Administration Guide 302-002-091 REV. 02 Copyright © 2010-2016 EMC Corporation. All rights reserved. Published in the USA.
Published March, 2016
EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without notice.
The information in this publication is provided as is. EMC Corporation makes no representations or warranties of any kind with respect to the information in this publication, and specifically disclaims implied warranties of merchantability or fitness for a particular purpose. Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.
EMC², EMC, and the EMC logo are registered trademarks or trademarks of EMC Corporation in the United States and other countries. All other trademarks used herein are the property of their respective owners.
For the most up-to-date regulatory document for your product line, go to EMC Online Support (https://support.emc.com).
EMC Corporation Hopkinton, Massachusetts 01748-9103 1-508-435-1000 In North America 1-866-464-7381 www.EMC.com
2 EMC Data Domain Operating System 5.7 Administration Guide CONTENTS
Chapter 1 EMC Data Domain System Features and Integration 17 Revision history. 18 EMC Data Domain system overview. 19 EMC Data Domain system features. 20 Data integrity. 20 Data deduplication. 20 Restore operations. 21 EMC Data Domain Replicator. 21 Multipath and load balancing. 21 High Availability. 21 System administrator access. 23 Licensed features. 23 Storage environment integration. 25
Chapter 2 Getting Started 29 DD System Manager overview. 30 Logging in and out of DD System Manager. 30 The DD System Manager interface. 31 Page elements. 32 Banner. 32 Navigation panel. 33 Information panel. 33 HA panel. 33 Footer. 33 Help buttons. 34 End User License Agreement. 34 Configuring the system with the configuration wizard. 34 License page. 35 Add License Key page. 35 Network General page. 35 Network Interfaces page. 36 Network DNS page. 36 System Settings Administrator page. 37 System Settings Email/Location page. 38 CIFS Protocol Authentication page. 38 CIFS Protocol Share page. 38 NFS Protocol Export page. 39 DD Boost Protocol Storage Unit page. 39 DD Boost Protocol Fibre Channel page. 40 VTL Protocol Library page. 40 VTL Protocol Access Group page. 41 Data Domain Command Line Interface. 42 Logging into the CLI. 42 CLI online help guidelines. 43
EMC Data Domain Operating System 5.7 Administration Guide 3 CONTENTS
Chapter 3 Managing Data Domain Systems 45 System management overview. 46 HA system management overview. 46 HA system planned maintenance. 47 Rebooting a system. 47 Powering a system on or off . 47 System upgrade management. 48 Viewing upgrade packages on the system. 48 Obtaining and verifying upgrade packages. 48 Upgrading a Data Domain system. 49 Removing an upgrade package. 50 Managing system licenses. 50 HA system license management. 51 System storage management. 51 Viewing system storage information. 51 Physically locating an enclosure. 55 Physically locating a disk. 55 Configuring storage. 56 Network connection management. 56 HA system network connection management. 56 Network interface management. 57 General network settings management. 71 Network route management. 75 System passphrase management. 77 Setting the system passphrase. 77 Changing the system passphrase. 78 System access management. 79 Role-based access control. 79 Access management for IP protocols. 80 Local user account management. 87 Directory user and group management. 94 Configuring mail server settings. 101 Managing time and date settings. 101 Managing system properties. 102 SNMP management. 103 Viewing SNMP status and configuration. 103 Enabling and disabling SNMP. 105 Downloading the SNMP MIB. 105 Configuring SNMP properties. 105 SNMP V3 user management. 106 SNMP V2C community management. 107 SNMP trap host management. 109 Autosupport report management. 111 HA system autosupport and support bundle manageability. 111 Enabling and disabling autosupport reporting to EMC. 112 Reviewing generated autosupport reports. 112 Configuring the autosupport mailing list. 112 Support bundle management. 113 Generating a support bundle. 113 Viewing the support bundles list. 114 Alert notification management. 114 HA system alert notification management. 114 Viewing the notification group list. 115 Creating a notification group. 116 Managing the subscriber list for a group. 117
4 EMC Data Domain Operating System 5.7 Administration Guide CONTENTS
Modifying a notification group. 117 Deleting a notification group. 118 Resetting the notification group configuration. 118 Configuring the daily summary schedule and distribution list. 118 Enabling and disabling alert notification to EMC. 119 Testing the alerts email feature. 120 EMC Support delivery management. 120 Selecting standard email delivery to EMC. 121 Selecting and configuring ConnectEMC delivery. 121 Testing ConnectEMC operation. 122 Log file management. 122 Viewing log files in DD System Manager. 123 Displaying a log file in the CLI. 123 Learning more about log messages. 124 Saving a copy of log files. 125 Log message transmission to remote systems. 125 Remote system power management with IPMI. 127 IPMI and SOL limitations. 127 Adding and deleting IPMI users with DD System Manager. 128 Changing an IPMI user password. 128 Configuring an IPMI port. 129 Preparing for remote power management and console monitoring with the CLI. 130 Managing power with DD System Manager. 131 Managing power with the CLI. 132
Chapter 4 Monitoring Data Domain Systems 135 Viewing individual system status and identity information. 136 Dashboard Alerts area. 136 Dashboard File System area. 137 Dashboard Services area. 137 Dashboard HA Readiness area. 137 Dashboard Hardware area. 138 Maintenance System area. 138 Health Alerts panel. 138 Viewing and clearing current alerts. 139 Current Alerts tab. 139 Viewing the alerts history. 140 Alerts History tab. 140 Viewing hardware component status. 141 Fan status. 142 Temperature status. 142 Power supply status. 143 PCI slot status. 143 NVRAM status. 143 Viewing system statistics. 144 Performance statistics graphs. 144 Viewing active users. 145 History report management. 146 Types of reports. 146 Creating a report. 150 Viewing saved reports. 151 Printing saved reports. 151 Deleting saved reports. 151 Renaming saved reports. 151
EMC Data Domain Operating System 5.7 Administration Guide 5 CONTENTS
Viewing the Task Log. 152 Viewing the system High Availability status. 153 High Availability status. 153
Chapter 5 File System 155 File system overview. 156 How the file system stores data. 156 How the file system reports space usage. 156 How the file system uses compression . 156 How the file system implements data integrity. 157 How the file system reclaims storage space with file system cleaning . 158 Supported interfaces . 159 Supported backup software. 159 Data streams sent to a Data Domain system . 159 File system limitations. 160 Monitoring file system usage. 161 Accessing the file system view. 162 Managing file system operations. 169 Performing basic operations. 169 Performing cleaning. 172 Performing sanitization. 173 Modifying basic settings. 174 Fast copy operations. 177 Performing a fast copy operation. 177
Chapter 6 MTrees 179 MTrees overview. 180 MTree limits. 180 Quotas. 180 About the MTree panel. 181 About the summary view. 181 About the space usage view (MTrees). 185 About the daily written view (MTrees). 186 Monitoring MTree usage. 187 Understanding physical capacity measurement. 187 Managing MTree operations. 190 Creating an MTree. 190 Configure and enable/disable MTree quotas. 191 Deleting an MTree. 192 Undeleting an MTree. 192 Renaming an MTree. 193 Replicating a system with quotas to one without. 193
Chapter 7 Snapshots 195 Snapshots overview. 196 Monitoring snapshots and their schedules. 197 About the snapshots view. 197 Managing snapshots. 198 Creating a snapshot. 198 Modifying a snapshot expiration date. 199 Renaming a snapshot. 199 Expiring a snapshot. 199
6 EMC Data Domain Operating System 5.7 Administration Guide CONTENTS
Managing snapshot schedules. 200 Creating a snapshot schedule. 200 Modifying a snapshot schedule. 201 Deleting a snapshot schedule. 201 Recover data from a snapshot. 202
Chapter 8 CIFS 203 CIFS overview. 204 Configuring SMB signing. 204 Performing CIFS setup. 205 HA systems and CIFS. 205 Preparing clients for access to Data Domain systems. 205 Enabling CIFS services. 205 Naming the CIFS server. 206 Setting authentication parameters. 206 Disabling CIFS services. 207 Working with shares. 207 Creating shares on the Data Domain system. 207 Modifying a share on a Data Domain system. 209 Creating a share from an existing share. 209 Disabling a share on a Data Domain system. 210 Enabling a share on a Data Domain system. 210 Deleting a share on a Data Domain system. 210 Performing MMC administration. 210 Connecting to a Data Domain system from a CIFS client. 210 Displaying CIFS information . 212 Managing access control. 212 Accessing shares from a Windows client. 212 Providing domain users administrative access. 213 Allowing administrative access to a Data Domain system for domain users. 213 Restricting administrative access from Windows. 214 File access. 214 Monitoring CIFS operation. 217 Displaying CIFS status. 217 Display CIFS configuration. 217 Displaying CIFS statistics. 220 Performing CIFS troubleshooting. 220 Displaying clients current activity. 220 Setting the maximum open files on a connection. 221 Data Domain system clock. 221 Synchronizing from a Windows domain controller. 222 Synchronize from an NTP server. 222
Chapter 9 NFS 223 NFS overview. 224 HA systems and NFS. 224 Managing NFS client access to the Data Domain system. 225 Enabling NFS services. 225 Disabling NFS services. 225 Creating an export. 225 Modifying an export. 227 Creating an export from an existing export. 228 Deleting an export. 228
EMC Data Domain Operating System 5.7 Administration Guide 7 CONTENTS
Displaying NFS information. 228 Viewing NFS status. 228 Viewing NFS exports. 229 Viewing active NFS clients. 229 Integrating a DDR into a Kerberos domain. 229 Add and delete KDC servers after initial configuration. 231
Chapter 10 Storage Migration 233 Storage migration overview. 234 Migration planning considerations. 234 Viewing migration status. 235 Evaluating migration readiness. 236 Migrating storage using DD System Manager. 236 Storage migration dialog descriptions. 237 Select a Task dialog. 237 Select Existing Enclosures dialog. 237 Select New Enclosures dialog. 238 Review Migration Plan dialog. 238 Verify Migration Preconditions dialog. 238 Migration progress dialogs. 239 Migrating storage using the CLI. 240 CLI storage migration example. 241
Chapter 11 SCSI Target 247 SCSI Target overview. 248 Fibre Channel view. 249 Enabling NPIV. 249 Disabling NPIV. 251 Resources tab. 252 Access Groups tab. 258 Differences in FC link monitoring among DD OS versions. 259
Chapter 12 Working with DD Boost 261 About Data Domain Boost software. 262 Managing DD Boost with DD System Manager. 263 Specifying DD Boost user names. 263 Changing DD Boost user passwords. 264 Removing a DD Boost user name. 264 Enabling DD Boost. 264 Disabling DD Boost. 264 Viewing DD Boost storage units. 265 Creating a storage unit. 266 Viewing storage unit information. 267 Modifying a storage unit. 270 Renaming a storage unit. 270 Deleting a storage unit. 271 Undeleting a storage unit. 271 Selecting DD Boost options. 272 Managing certificates for DD Boost. 273 Managing DD Boost client access and encryption. 275 About interface groups. 276 Interfaces. 277 Clients. 278
8 EMC Data Domain Operating System 5.7 Administration Guide CONTENTS
Creating interface groups. 278 Enabling and disabling interface groups. 279 Modifying an interface group’s name and interfaces. 279 Deleting an interface group. 280 Adding a client to an interface group. 280 Modifying a client’s name or interface group. 281 Deleting a client from the interface group. 281 Using interface groups for Managed File Replication (MFR). 281 Destroying DD Boost. 283 Configuring DD Boost-over-Fibre Channel. 284 Enabling DD Boost users. 284 Configuring DD Boost. 285 Verifying connectivity and creating access groups. 286 Using DD Boost on HA systems. 288 About the DD Boost tabs. 288 Settings. 289 Active Connections. 289 IP Network. 290 Fibre Channel. 290 Storage Units. 290
Chapter 13 DD Virtual Tape Library 293 DD Virtual Tape Library overview. 294 Planning a VTL. 294 VTL limits. 295 Number of drives supported by a VTL. 297 Tape barcodes. 297 LTO tape drive compatibility. 298 Setting up a VTL. 299 HA systems and VTL. 299 Managing a VTL. 299 Enabling VTL. 301 Disabling VTL. 301 VTL option defaults. 301 Configuring VTL default options. 302 Working with libraries. 303 Creating libraries. 303 Deleting libraries. 305 Searching for tapes. 306 Working with a selected library. 306 Creating tapes. 307 Deleting tapes. 308 Importing tapes. 309 Exporting tapes. 311 Moving tapes between devices within a library. 311 Adding slots. 313 Deleting slots. 313 Adding CAPs. 313 Deleting CAPs. 314 Viewing changer information. 314 Working with drives. 315 Creating drives. 316 Deleting drives. 316 Working with a selected drive. 317 Working with tapes. 317
EMC Data Domain Operating System 5.7 Administration Guide 9 CONTENTS
Changing a tape’s write or retention lock state. 318 Working with the vault. 319 Working with access groups. 319 Creating an access group. 320 Deleting an access group. 323 Working with a selected access group. 323 Selecting endpoints for a device. 324 Configuring the NDMP device TapeServer group. 324 Working with resources. 325 Working with initiators. 326 Working with endpoints. 327 Working with a selected endpoint. 328 Working with pools. 330 Creating pools. 331 Deleting pools. 332 Working with a selected pool. 332 Converting a directory pool to an MTree pool . 333 Moving tapes between pools. 334 Copying tapes between pools. 335 Renaming pools. 336
Chapter 14 DD Replicator 337 DD Replicator overview. 338 Replication between HA and non-HA systems. 338 Prerequisites for replication configuration. 339 Replication version compatibility. 340 Using DD Encryption with DD Replicator. 343 Replication types. 344 Managed file replication . 345 Directory replication. 345 MTree replication. 346 Collection replication . 347 Replication topologies. 348 One-to-one replication. 349 Bi-directional replication. 350 One-to-many replication. 351 Many-to-one replication. 351 Cascaded replication. 352 Managing replication. 353 Replication status. 354 Summary view. 354 DD Boost view. 364 Topology view. 365 Performance view. 366 Advanced Settings view. 366 Monitoring replication . 369 Checking replication pair status. 369 Viewing estimated completion time for backup jobs. 369 Checking replication context performance. 370 Tracking status of a replication process. 370
Chapter 15 DD Secure Multitenancy 371 Data Domain Secure Multitenancy overview. 372 SMT architecture basics. 372
10 EMC Data Domain Operating System 5.7 Administration Guide CONTENTS
Terminology used in SMT. 372 Control path and administrative isolation. 373 Understanding RBAC in SMT. 374 Provisioning a Tenant Unit. 375 Enabling Tenant Self-Service mode. 378 Data access by protocol. 379 Multi-User DD Boost and Storage Units in SMT. 379 Configuring access for CIFS. 379 Configuring NFS access. 380 Configuring access for VTL. 380 Using VTL NDMP TapeServer . 381 Data management operations. 381 Collecting performance statistics. 381 Modifying quotas. 381 SMT and replication. 382 SMT Tenant alerts. 383 Managing snapshots. 383 Performing a file system Fast Copy. 383
Chapter 16 DD Extended Retention 385 DD Extended Retention overview. 386 Supported protocols in DD Extended Retention. 388 High Availability and Extended Retention. 388 Using DD Replicator with DD Extended Retention. 388 Collection replication with DD Extended Retention. 388 Directory replication with DD Extended Retention. 389 MTree replication with DD Extended Retention. 389 Managed file replication with DD Extended Retention. 389 Hardware and licensing for DD Extended Retention. 390 Hardware supported for DD Extended Retention. 390 Licensing for DD Extended Retention. 392 Adding shelf capacity licenses for DD Extended Retention. 392 Configuring storage for DD Extended Retention. 393 Customer-provided infrastructure for DD Extended Retention. 393 Managing DD Extended Retention. 393 Enabling DD systems for DD Extended Retention. 394 Creating a two-tiered file system for DD Extended Retention. 394 File system panel for DD Extended Retention. 395 File system tabs for DD Extended Retention. 397 Upgrades and recovery with DD Extended Retention. 403 Upgrading to DD OS 5.7 with DD Extended Retention. 403 Upgrading hardware with DD Extended Retention. 403 Recovering a DD Extended Retention-enabled system. 404
Chapter 17 DD Retention Lock 405 DD Retention Lock overview. 406 DD Retention Lock protocol. 407 DD Retention Lock flow. 407 Supported data access protocols. 408 Enabling DD Retention Lock on an MTree. 409 Enabling DD Retention Lock Governance on an MTree. 409 Enabling DD Retention Lock Compliance on an MTree. 410 Client-Side Retention Lock file control. 412 Setting Retention Locking on a file. 413
EMC Data Domain Operating System 5.7 Administration Guide 11 CONTENTS
Extending Retention Locking on a file. 415 Identifying a Retention-Locked file. 416 Specifying a directory and touching only those files. 416 Reading a list of files and touching only those files. 416 Deleting or expiring a file. 417 Using ctime or mtime on Retention-Locked files. 417 System behavior with DD Retention Lock. 417 DD Retention Lock governance. 418 DD Retention Lock compliance. 419
Chapter 18 DD Encryption 431 DD encryption overview. 432 Configuring encryption. 433 About key management. 434 Rectifying lost or corrupted keys. 434 Key manager support. 435 Working with the RSA DPM Key Manager. 435 Working with the Embedded Key Manager. 437 How the cleaning operation works. 438 Key manager setup. 438 RSA DPM Key Manager encryption setup. 438 Changing key managers after setup. 441 Managing certificates for RSA Key Manager. 442 Checking settings for encryption of data at rest. 443 Enabling and disabling encryption of data at rest. 443 Enabling encryption of data at rest. 443 Disabling encryption of data at rest. 443 Locking and unlocking the file system. 444 Locking the file system. 444 Unlocking the file system. 445 Changing the encryption algorithm. 445
12 EMC Data Domain Operating System 5.7 Administration Guide Preface
As part of an effort to improve its product lines, EMC periodically releases revisions of its software and hardware. Therefore, some functions described in this document might not be supported by all versions of the software or hardware currently in use. The product release notes provide the most up-to-date information on product features. Contact your EMC technical support professional if a product does not function properly or does not function as described in this document.
This document was accurate at publication time. Go to EMC Online Support (https:// support.emc.com) to ensure that you are using the latest version of this document.
Purpose This guide explains how to manage the EMC Data Domain® systems with an emphasis on procedures using the EMC Data Domain System Manager (DD System Manager), a browser-based graphical user interface (GUI). If an important administrative task is not supported in DD System Manager, the Command Line Interface (CLI) commands are described.
l DD System Manager was formerly known as the Enterprise Manager. l In some cases, a CLI command may offer more options than those offered by the corresponding DD System Manager feature. See the EMC Data Domain Operating System Command Reference Guide for a complete description of a command and its options.
Audience This guide is for system administrators who are familiar with standard backup software packages and general backup administration. Related documentation The following Data Domain system documents provide additional information:
l Installation and setup guide for your system, for example, EMC Data Domain DD2500 Storage System, Installation and Setup Guide l EMC Data Domain DD9500 High Availability System Installation and Upgrade Guide l EMC Data Domain Operating System USB Installation Guide l EMC Data Domain Operating System DVD Installation Guide l EMC Data Domain Operating System Release Notes l EMC Data Domain Operating System Initial Configuration Guide l EMC Data Domain Product Security Guide l EMC Data Domain Operating System High Availability White Paper l EMC Data Domain Operating System Command Reference Guide l EMC Data Domain Operating System MIB Quick Reference
Preface 13 Preface
l EMC Data Domain Operating System Offline Diagnostics Suite User's Guide l Hardware overview guide for your system, for example, EMC Data Domain DD4200, DD4500, and DD7200 Systems, Hardware Overview l Field replacement guides for your system components, for example, Field Replacement Guide, Data Domain DD4200, DD4500, and DD7200 Systems, IO Module and Management Module Replacement or Upgrade l EMC Data Domain, System Controller Upgrade Guide l EMC Data Domain Expansion Shelf, Hardware Guide (for shelf model ES20 or ES30) l EMC Data Domain Boost for Partner Integration Administration Guide l EMC Data Domain Boost for OpenStorage Administration Guide l EMC Data Domain Boost for Oracle Recovery Manager Administration Guide l EMC Data Domain Boost SDK Programmer's Guide l Statement of Volatility for the Data Domain DD2500 System l Statement of Volatility for the Data Domain DD4200, DD4500, or DD7200 System If you have the optional RSA Data Protection (DPM) Key Manager, see the latest version of the RSA Data Protection Manager Server Administrator's Guide, available with the RSA Key Manager product. Special notice conventions used in this document EMC uses the following conventions for special notices:
A notice identifies content that warns of a potential business or data loss.
A note identifies information that is incidental, but not essential, to the topic. Notes can provide an explanation, a comment, reinforcement of a point in the text, or just a related point.
Typographical conventions EMC uses the following type style conventions in this document:
Table 1 Typography
Bold Indicates interface element names, such as names of windows, dialog boxes, buttons, fields, tab names, key names, and menu paths (what the user specifically selects or clicks)
Italic Highlights publication titles listed in text Monospace Indicates system information, such as:
l System output, such as an error message or script
l Pathnames, filenames, prompts, and syntax
l Commands and options
Monospace italic Highlights a variable name that must be replaced with a variable value Monospace bold Indicates text for user input [ ] Square brackets enclose optional values
14 EMC Data Domain Operating System 5.7 Administration Guide Preface
Table 1 Typography (continued)
| Vertical bar indicates alternate selections—the bar means “or” < >Braces enclose content that the user must specify, such as x or y or z . Ellipses indicate nonessential information omitted from the example
Where to get help The following topics describe how to get more product information and contact technical support. Product information For documentation, release notes, software updates, or information about EMC products, go to EMC Online Support at https://support.emc.com. EMC Data Domain product documentation To view documentation for EMC Data Domain products, go to EMC Online Support and click Support by Product below the Search box. Type Data Domain in the Find a Product box, wait for those words to appear in the list of matches below the box, and click the words. Then click >>. In the list of categories under the Search box, click Documentation.
l The Product choices let you filter results by Data Domain system model number, such as DD990, or by DD OS software release.
l The Content Type choices let you filter results by category. Click More under Content Type to see all of the categories. The categories that contain end-user and compatibility documentation are:
n Manuals and Guides, for the software and hardware manuals for your system, and for integration guides that explain how to use EMC Data Domain systems with backup software and other products
n Release Notes, for specific versions of the EMC Data Domain Operating System and EMC Data Domain products
n Compatibility Document, for guides that show which EMC and third-party components are compatible
Technical support Go to EMC Online Support and click Service Center. You will see several options for contacting EMC Technical Support. Note that to open a service request, you must have a valid support agreement. Contact your EMC sales representative for details about obtaining a valid support agreement or with questions about your account.
Your comments Your suggestions will help us continue to improve the accuracy, organization, and overall quality of the user publications. Send your opinions of this document to: [email protected].
16 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 1 EMC Data Domain System Features and Integration
This chapter includes:
l Revision history. 18 l EMC Data Domain system overview. 19 l EMC Data Domain system features. 20 l Storage environment integration. 25
EMC Data Domain System Features and Integration 17 EMC Data Domain System Features and Integration
Revision history The revision history lists the major changes to this document to support DD OS Release 5.7.
Table 2 Document revision history
Revision Date Description
02 (5.7.1) March 2016 This revision includes information about the Data Domain High Availability (HA) feature, including:
l An overview of the HA feature and its architecture
l HA features in the System Manager user interface
l HA system management information, including planned maintenance, licensing, networking, autosupport, and alerts management
l Information about viewing HA status
l Considerations for using DD Boost, CIFS, NFS, and VTL with HA systems
l Information about replication between HA and non-HA systems
01 (5.7.0) October 2015 This revision includes information about these new features:
l The new FM 2.0/Likewise server functionality enables faster recall rates to support the active archive, and the system now supports both SMB 1.0 and 2.1
l The new physical capacity measurement feature provides space usage information for a sub-set of storage space
l You can now configure ConnectEMC delivery with DD System Manager
l The new storage migration feature allows you to transfer all files from one set of enclosures to another set
l DD System Manager supports the new DS60 enclosure and 4 TB drives in DS60 and ES30 enclosures
l A new option allows you to force new users to change their password during their first login
l The new MTree scaling feature pushes MTree support from 32 to up to 128 or 256 (DD9500 only) active mtrees depending upon the platform
l The new feature regarding 2000 write streams for SMT on Apollo
l The privileged delete feature enables deletion of Retention Lock Governance files SCSI Target changes:
18 EMC Data Domain Operating System 5.7 Administration Guide EMC Data Domain System Features and Integration
Table 2 Document revision history (continued)
Revision Date Description
l NPIV (N-port ID Virtualization), a Fibre Channel feature that lets multiple endpoints share a single physical port, eases hardware requirements, and provides failover capabilities, is now supported DD Boost changes:
l DFC (DD Boost-over-Fibre Channel) now supports the Solaris client environment
l The DD Boost Active Connections graph provides statistics for active DD Boost connections
l Interface groups can be used for Managed File Replication
l A special license, BLOCK-SERVICES-PROTECTPOINT, is available to enable clients using ProtectPoint block services to have DD Boost functionality without a DD Boost license VTL changes:
l Barcode lengths of 6 or 8 characters are now supported, depending on the changer model type Removed features:
l Data Domain System Manager no longer manages multiple systems. To manage multiple systems from a single program, use Data Domain Management Center
l The following system models are not supported in this release: DD660, DD690, and DD880
EMC Data Domain system overview EMC Data Domain systems are disk-based inline deduplication appliances that provide data protection and disaster recovery (DR) in the enterprise environment. All systems run the EMC Data Domain Operating System (DD OS), which provides both a command-line interface (CLI) for performing all system operations, and the EMC Data Domain System Manager (DD System Manager) graphical user interface (GUI) for configuration, management, and monitoring.
DD System Manager was formerly known as the Enterprise Manager.
Systems consist of appliances that vary in storage capacity and data throughput. Systems are typically configured with expansion enclosures that add storage space.
EMC Data Domain system overview 19 EMC Data Domain System Features and Integration
EMC Data Domain system features Data Domain system features ensure data integrity, reliable restoration, efficient resource usage, and ease of management. Licensed features allow you to scale the system feature set to match your needs and budget.
Data integrity The DD OS Data Invulnerability Architecture™ protects against data loss from hardware and software failures. l When writing to disk, the DD OS creates and stores checksums and self-describing metadata for all data received. After writing the data to disk, the DD OS then recomputes and verifies the checksums and metadata. l An append-only write policy guards against overwriting valid data. l After a backup completes, a validation process examines what was written to disk and verifies that all file segments are logically correct within the file system and that the data is identical before and after writing to disk. l In the background, the online verify operation continuously checks that data on the disks is correct and unchanged since the earlier validation process. l Storage in most Data Domain systems is set up in a double parity RAID 6 configuration (two parity drives). Additionally, most configurations include a hot spare in each enclosure, except the DD1xx series systems, which use eight disks. Each parity stripe uses block checksums to ensure that data is correct. Checksums are constantly used during the online verify operation and while data is read from the Data Domain system. With double parity, the system can fix simultaneous errors on as many as two disks. l To keep data synchronized during a hardware or power failure, the Data Domain system uses NVRAM (non-volatile RAM) to track outstanding I/O operations. An NVRAM card with fully charged batteries (the typical state) can retain data for a period of hours, which is determined by the hardware in use. l When reading data back on a restore operation, the DD OS uses multiple layers of consistency checks to verify that restored data is correct.
Data deduplication DD OS data deduplication identifies redundant data during each backup and stores unique data just once. The storage of unique data is invisible to backup software and independent of data format. Data can be structured, such as databases, or unstructured, such as text files. Data can derive from file systems or from raw volumes. Typical deduplication ratios are 20-to-1, on average, over many weeks. This ratio assumes there are weekly full backups and daily incremental backups. A backup that includes many duplicate or similar files (files copied several times with minor changes) benefits the most from deduplication. Depending on backup volume, size, retention period, and rate of change, the amount of deduplication can vary. The best deduplication happens with backup volume sizes of at least 10 MiB (MiB is the base 2 equivalent of MB). To take full advantage of multiple Data Domain systems, a site with more than one Data Domain system must consistently backup the same client system or set of data to the same Data Domain system. For example, if a full back up of all sales data goes to Data
20 EMC Data Domain Operating System 5.7 Administration Guide EMC Data Domain System Features and Integration
Domain system A, maximum deduplication is achieved when the incremental backups and future full backups for sales data also go to Data Domain system A.
Restore operations File restore operations create little or no contention with backup or other restore operations. When backing up to disks on a Data Domain system, incremental backups are always reliable and can be easily accessed. With tape backups, a restore operation may rely on multiple tapes holding incremental backups. Also, the more incremental backups a site stores on multiple tapes, the more time-consuming and risky the restore process. One bad tape can kill the restore. Using a Data Domain system, you can perform full backups more frequently without the penalty of storing redundant data. Unlike tape drive backups, multiple processes can access a Data Domain system simultaneously. A Data Domain system allows your site to offer safe, user-driven, single-file restore operations.
EMC Data Domain Replicator The EMC Data Domain Replicator sets up and manages the replication of backup data between two Data Domain systems. A Replicator pair consists of a source and a destination system and replicates a complete data set or directory from the source system to the destination system. An individual Data Domain system can be a part of multiple replication pairs and can serve as a source for one or more pairs and a destination for one or more pairs. After replication is started, the source system automatically sends any new backup data to the destination system.
Multipath and load balancing In a Fibre Channel multipath configuration, multiple paths are established between a Data Domain system and a backup server or backup destination array. When multiple paths are present, the system automatically balances the backup load between the available paths. At least two HBA ports are required to create a multipath configuration. When connected to a backup server, each of the HBA ports on the multipath is connected to a separate port on the backup server.
High Availability The High Availability (HA) feature lets you configure two DD9500 systems as an Active- Standby pair, providing redundancy in the event of a system failure. HA keeps the active and standby systems in sync, so that if the active node were to fail due to hardware or software issues, the standby node can take over services and continue where the failing node left off. The HA feature:
l Supports failover of backup, restore, replication and management services in a two- node system. Automatic failover requires no user intervention. l Provides a fully redundant design with no single point of failure within the system when configured as recommended. l Provides an Active-Standby system with no loss of performance on failover. l Provides failover within 10 minutes for most operations. CIFS, VTL, and NDMP must be restarted manually.
Restore operations 21 EMC Data Domain System Features and Integration
Recovery of DD Boost applications may take longer than 10 minutes, because Boost application recovery cannot begin until the DD server failover is complete. In addition, Boost application recovery cannot start until the application invokes the Boost library. Similarly, NFS may require additional time to recover.
l Supports ease of management and configuration through DD OS CLIs. l Provides alerts for malfunctioning hardware. l Preserves single-node performance and scalability within an HA configuration in both normal and degraded mode. l Supports the same feature set as stand-alone DD systems.
Extended Retention and vDisk are not supported.
l Supports DD9500 systems with all SAS drives. This includes legacy systems upgraded to DD9500 systems with all SAS drives.
The Data Domain DD9500 High Availability System Installation and Upgrade Guide describes how to install a new DD9500 HA system, or upgrade an existing DD9500 to an HA pair.
l Does not impact the ability to scale the product. l Supports nondisruptive software updates.
HA architecture HA functionality is available for both IP and FC connections. Both nodes must have access to the same IP networks, FC SANs, and hosts in order to achieve high availability for the environment. Over IP networks, HA uses a floating IP address to provide data access to the Data Domain HA pair regardless of which physical node is the active node. Over FC SANs, HA uses NPIV to move the FC WWNs between nodes, allowing the FC initiators to re-establish connections after a failover. Figure 1 on page 23 shows the HA architecture.
22 EMC Data Domain Operating System 5.7 Administration Guide EMC Data Domain System Features and Integration
Figure 1 HA architecture
System administrator access System administrators can access the system for configuration and management using a command line interface or a graphical user interface.
l DD OS CLI—A command-line interface that is available through a serial console or through Ethernet connections using SSH or Telnet. CLI commands enable initial system configuration, changes to individual system settings, and display of system operation status. l DD System Manager—A browser-based graphical user interface that is available through Ethernet connections. Use DD System Manager to perform initial system configuration, make configuration changes after initial configuration, display system and component status, and generate reports and charts.
Some systems support access using a keyboard and monitor attached directly to the system.
Licensed features Feature licenses allow you to purchase only those features you intend to use. Some examples of features that require licenses are Extended Retention, DD Boost, and storage capacity increases. Consult with your EMC representative for information on purchasing licensed features.
System administrator access 23 EMC Data Domain System Features and Integration
Table 3 Features requiring licenses
Feature Name License Name in Description Software
EMC Data Domain ARCHIVESTORE Licenses Data Domain systems for archive use, ArchiveStore such as file and email archiving, file tiering, and content and database archiving. EMC Data Domain DDBOOST Enables the use of a Data Domain system with the Boost following applications: EMC Avamar, EMC NetWorker, Oracle RMAN, Quest vRanger, Symantec Veritas NetBackup (NBU), and Backup Exec. The managed file replication (MFR) feature of DD Boost also requires the DD Replicator license. EMC Data Domain CONTROLLER-COD Enables an on-demand capacity increase for 4 TB Capacity on Demand DD2200 systems to 7.5 TB or 13.18 TB. An increase to 13.18 TB also requires the EXPANDED-STORAGE license.
EMC Data Domain ENCRYPTION Allows data on system drives or external storage to Encryption be encrypted while being saved and locked when moving the system to another location. EMC Data Domain EXPANDED- Allows Data Domain system storage to be Expansion Storage STORAGE expanded beyond the level provided in the base system. EMC Data Domain EXTENDED- Licenses the Extended Retention storage feature. Extended Retention RETENTION (formerly DD Archiver) EMC Data Domain I/OS I/OS An I/OS license is required when VTL is used to (for IBM i operating backup systems in the IBM i operating environments) environment. Apply this license before adding virtual tape drives to libraries. EMC Data Domain REPLICATION Adds DD Replicator for replication of data from one Replicator Data Domain system to another. A license is required on each system. EMC Data Domain RETENTION-LOCK- Meets the strictest data retention requirements Retention Lock COMPLIANCE from regulatory standards such as SEC17a-4. Compliance Edition EMC Data Domain RETENTION-LOCK- Protects selected files from modification and Retention Lock GOVERNANCE deletion before a specified retention period Governance Edition expires. EMC Data Domain CAPACITY-ACTIVE Enables a Data Domain system to expand the Shelf Capacity-Active active tier storage capacity to an additional Tier enclosure or a disk pack within an enclosure. EMC Data Domain CAPACITY-ARCHIVE Enables a Data Domain system to expand the Shelf Capacity-Archive archive tier storage capacity to an additional Tier enclosure or a disk pack within an enclosure. EMC Data Domain STORAGE- Enables migration of data from one enclosure to Storage Migration MIGRATION-FOR- another to support replacement of older, lower- capacity enclosures.
24 EMC Data Domain Operating System 5.7 Administration Guide EMC Data Domain System Features and Integration
Table 3 Features requiring licenses (continued)
Feature Name License Name in Description Software
DATADOMAIN- SYSTEMS EMC Data Domain VTL Enables the use of a Data Domain system as a Virtual Tape Library virtual tape library over a Fibre Channel network. (VTL) This license also enables the NDMP Tape Server feature, which previously required a separate license. EMC High Availability HA-ACTIVE-PASSIVE Enables the High Availability feature in an Active- Standby configuration. You only need to purchase one HA license; the license runs on the active node and is mirrored to the standby node.
Storage environment integration EMC Data Domain systems integrate easily into existing data centers.
l All Data Domain systems can be configured as storage destinations for leading backup and archiving applications using NFS, CIFS, Boost, or VTL protocols. l Search for compatibility documents at https://support.emc.com for information on the applications that work with the different configurations. l Multiple backup servers can share one Data Domain system. l One Data Domain system can handle multiple simultaneous backup and restore operations. l Multiple Data Domain systems can be connected to one or more backup servers. For use as a backup destination, a Data Domain system can be configured either as a disk storage unit with a file system that is accessed through an Ethernet connection or as a virtual tape library (VTL) that is accessed through a Fibre Channel connection. The VTL feature enables Data Domain systems to be integrated into environments where backup software is already configured for tape backups, minimizing disruption. Configuration is performed both in the DD OS, as described in the relevant sections of this guide, and in the backup application, as described in the backup application’s administrator guides and in Data Domain application-related guides and tech notes.
l All backup applications can access a Data Domain system as either an NFS or a CIFS file system on the Data Domain disk device. l The following applications work with a Data Domain system using the DD Boost interface: EMC Avamar, EMC NetWorker, Oracle RMAN, Quest vRanger, Symantec Veritas NetBackup (NBU), and Backup Exec. The following figure shows a Data Domain system integrated into an existing basic backup configuration.
Storage environment integration 25 EMC Data Domain System Features and Integration
Figure 2 Data Domain system integrated into a storage environment
1. Primary storage 2. Ethernet 3. Backup server 4. SCSI/Fibre Channel 5. Gigabit Ethernet or Fibre Channel 6. Tape system 7. Data Domain system 8. Management 9. NFS/CIFS/VTL/DD Boost 10. Data Verification 11. Data Domain file system 12. Global deduplication and compression 13. RAID
As shown in Figure 2 on page 26, data flows to a Data Domain system through an Ethernet or Fibre Channel connection. Immediately, the data verification processes begin and are continued while the data resides on the Data Domain system. In the file system, the DD OS Global Compression™ algorithms dedupe and compress the data for storage. Data is then sent to the disk RAID subsystem. When a restore operation is required, data is retrieved from Data Domain storage, decompressed, verified for consistency, and
26 EMC Data Domain Operating System 5.7 Administration Guide EMC Data Domain System Features and Integration transferred via Ethernet to the backup servers using Ethernet (for NFS, CIFS, DD Boost), or using Fiber Channel (for VTL and DD Boost). The DD OS accommodates relatively large streams of sequential data from backup software and is optimized for high throughput, continuous data verification, and high compression. It also accommodates the large numbers of smaller files in nearline storage (DD ArchiveStore). Data Domain system performance is best when storing data from applications that are not specifically backup software under the following circumstances. l Data is sent to the Data Domain system as sequential writes (no overwrites). l Data is neither compressed nor encrypted before being sent to the Data Domain system.
Storage environment integration 27 EMC Data Domain System Features and Integration
28 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 2 Getting Started
This chapter includes:
l DD System Manager overview. 30 l Logging in and out of DD System Manager. 30 l The DD System Manager interface. 31 l Configuring the system with the configuration wizard. 34 l Data Domain Command Line Interface. 42 l Logging into the CLI. 42 l CLI online help guidelines. 43
Getting Started 29 Getting Started
DD System Manager overview DD System Manager is a browser-based graphical user interface, available through Ethernet connections, for managing a single system from any location. DD System Manager provides a single, consolidated management interface that allows for configuration and monitoring of many system features and system settings.
The Data Domain Management Center allows you to manage multiple systems from a single browser window.
DD System Manager provides real-time graphs and tables that allow you to monitor the status of system hardware components and configured features. Additionally, a command set that performs all system functions is available to users at the command-line interface (CLI). Commands configure system settings and provide displays of system hardware status, feature configuration, and operation. The command-line interface is available through a serial console or through an Ethernet connection using SSH or Telnet.
Some systems support access using a keyboard and monitor attached directly to the system.
Logging in and out of DD System Manager Use a browser to log in to DD System Manager. Procedure 1. Open a web browser and enter the IP address or hostname to connect to DD System Manager. It must be: l A fully qualified domain name (for example, http://dd01.emc.com) l A hostname (http://dd01) l An IP address (http://10.5.50.5)
DD System Manager uses HTTP port 80 and HTTPS port 443. If your Data Domain system is behind a firewall, you may need to enable port 80 if using HTTP, or port 443 if using HTTPS to reach the system. The port numbers can be easily changed if security requirements dictate.
2. For HTTPS secure login, click Secure Login. Secure login with HTTPS requires a digital certificate to validate the identity of the DD OS system and to support bidirectional encryption between DD System Manager and a browser. DD OS includes a self-signed certificate, and DD OS allows you to import your own certificate. The default settings of most browsers do not automatically accept a self-signed certificate. This does not prevent you from using the self-signed certificate; it just means that you must respond to a warning message each time you perform a secure
30 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
log in, or you must install the certificate in your browser. For instructions on how to install the certificate in your browser, see your browser documentation. 3. Enter your assigned username and password.
The initial username is sysadmin and the initial password is the system serial number. For information on setting up a new system, see the EMC Data Domain Operating System Initial Configuration Guide.
4. Click Log In. If this is the first time you have logged in, the Home view appears in the Information panel.
If you enter an incorrect password 4 consecutive times, the system locks out the specified username for 120 seconds. The login count and lockout period are configurable and might be different on your system.
If this is the first time you are logging in, you might be required to change your password. If the system administrator has configured your username to require a password change, you must change the password before gaining access to DD System Manager.
5. To log out, click the log out button in the DD System Manager banner. When you log out, the system displays the log in page with a message that your log out is complete.
The DD System Manager interface The DD System Manager interface provides common elements on most pages that enable you to navigate through the configuration and display options and display context sensitive help.
The DD System Manager interface 31 Getting Started
Page elements The primary page elements are the banner, the navigation panel, the information panels, and footer. Figure 3 DD System Manager page components
1. Banner 2. Navigation panel 3. Information panels 4. Footer
Banner The DD System Manager banner displays the program name and buttons for Refresh, Log Out, and Help.
32 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
Navigation panel The Navigation panel displays the highest level menu selections that you can use to identify the system component or task that you want to manage. The Navigation panel displays the top two levels of the navigation system. Click any top level title to display the second level titles. Tabs and menus in the Information panel provide additional navigation controls.
Information panel The Information panel displays information and controls related to the selected item in the Navigation panel. The information panel is where you find system status information and configure a system. Depending on the feature or task selected in the Navigation panel, the Information panel may display a tab bar, topic areas, table view controls, and the More Tasks menu. Tab bar Tabs provide access to different aspects of the topic selected in the Navigation panel. Topic areas Topic areas divide the Information panel into sections that represent different aspects of the topic selected in the Navigation panel or parent tab. Working with table view options Many of the views with tables of items contain controls for filtering, navigating, and sorting the information in the table. How to use common table controls:
l Click the diamond icon in a column heading to reverse the sort order of items in the column. l Click the arrows at the bottom right of the view to move forward or backward through the pages. To skip to the beginning of a sequence of pages, click | |. l Use the scroll bar to view all items in a table. l Enter text in the Filter By box to search for or prioritize the listing of those items. l Click Update to refresh the list. l Click Reset to return to the default listing. More Tasks menu Some pages provide a More Tasks menu at the top right of the view that contains commands related to the current view.
HA panel In high-availability (HA) systems, the HA panel indicates whether the system can fail over from the active node to the standby node if necessary. You can click on the HA panel to navigate to the HA section under ADMINISTRATION.
Footer The DD System Manager footer displays important information about the management session. The banner lists the following information.
Navigation panel 33 Getting Started
l System hostname. l DD OS version l Selected system model number. l User name and role for the current logged in user.
Help buttons Help buttons display a ? and appear in the banner, in the title of many areas of the Information panel, and in many dialogs. Click the help button to display a help window related to the current feature you are using. The help window provides a contents button and navigation button above the help. Click the contents button to display the guide contents and a search button that you can use to search the help. Use the directional arrow buttons to page through the help topics in sequential order.
End User License Agreement To view the End User License Agreement (EULA), select Maintenance > System > More Tasks > EULA.
Configuring the system with the configuration wizard There are two wizards, a DD System Manager configuration wizard and a Command Line Interface (CLI) configuration wizard. The configuration wizards guide you through a simplified configuration of your system to get your system operating quickly. After you complete the basic configuration with a wizard, you can use additional configuration controls in DD System Manager and the CLI to further configure your system.
The following procedure describes how to start and run the DD System Manager configuration wizard after the initial configuration of your system. For instructions on running the configuration wizards at system startup, see the EMC Data Domain Operating System Initial Configuration Guide.
If you want to configure your system for high availability (HA), you must perform this operation using the CLI Configuration Wizard. For more information, see the EMC Data Domain DD9500 System Version 5.7.1 High Availability (HA) System Installation and Upgrade Guide and the EMC Data Domain Operating System Version 5.7.1 Initial Configuration Guide.
1. Select Maintenance > System > More Tasks > Launch Configuration Wizard. 2. Use the controls at the bottom of the Configuration Wizard dialog to select which features you want to configure and to advance through the wizard. To display help for a feature, click the help icon (question mark) in the lower left corner of the dialog.
34 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
License page The License page displays all installed licenses. Click Yes to add a license or click No to skip license installation.
Add License Key page The Add License Key page accepts a single license per line. Press Enter after you enter each license key, and click Add when you are done.
Network General page The General page allows you to configure network settings that define how the system participates in an IP network. To configure these network settings outside of the configuration wizard, select Hardware > Ethernet.
Table 4 General page settings
Obtain Settings using DHCP Select this option to specify that the system collect network settings from a Dynamic Host Control Protocol (DHCP) server. When you configure the network interfaces, at least one of the interfaces must be configured to use DHCP. Manually Configure Select this option to use the network settings defined in the Settings area of this page. Host Name Specifies the network hostname for this system.
If you choose to obtain the network settings through DHCP, you can manually configure the hostname at Hardware > Ethernet > Settings or with the net set hostname command. You must manually configure the host name when using DHCP over IPv6.
Domain Name Specifies the network domain to which this system belongs. Default IPv4 Gateway Specifies the IPv4 address of the gateway to which the system will forward network requests when there is no route entry for the destination system.
Default IPv6 Gateway Specifies the IPv6 address of the gateway to which the system will forward network requests when there is no route entry for the destination system.
License page 35 Getting Started
Network Interfaces page The Interfaces page allows you to configure network settings that define how each interface participates in an IP network. To Configure these network settings outside of the configuration wizard, select Hardware > Ethernet > Interfaces.
Table 5 Interfaces page settings
Interface Lists the interfaces available on your system. Enabled Shows whether each interface is enabled (checkbox selected) or disabled (not selected). Click the checkbox to toggle the interface between the enabled and disabled states.
DHCP Shows the current Dynamic Host Control Protocol (DHCP) configuration for each interface. Select v4 for IPv4 DHCP connections, v6 for IPv6 connections, or no to disable DHCP. IP Address Specifies an IPv4 or IPv6 address for this system. To configure the IP address, you must set DHCP to No.
DD140, DD160, DD610, DD620, and DD630 systems do not support IPv6 on interface eth0a (eth0 on systems that use legacy port names) or on any VLANs created on that interface.
Netmask Specifies the network mask for this system. To configure the network mask, you must set DHCP to No. Link Displays whether the Ethernet link is active (Yes) or not (No).
Network DNS page The DNS page allows you to configure how the system obtains IP addresses for DNS servers in a Domain Name System (DNS). To Configure these network settings outside of the configuration wizard, select Hardware > Ethernet > Settings.
Table 6 DNS page settings
Obtain DNS using DHCP. Select this option to specify that the system collect DNS IP addresses from a Dynamic Host Control Protocol (DHCP) server. When you configure the network interfaces, at least one of the interfaces must be configured to use DHCP. Manually configure DNS list. Select this option when you want to manually enter DNS server IP addresses. Add (+) button Click this button to display a dialog in which you can add a DNS IP address to the DNS IP Address list. You must select
36 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
Table 6 DNS page settings (continued)
Manually configure DNS list before you can add or delete DNS IP addresses. Delete (X) button Click this button to delete a DNS IP address from the DNS IP Address list. You must select the IP address to delete before this button is enabled. You must also select Manually configure DNS list before you can add or delete DNS IP addresses. IP Address Checkboxes Select a checkbox for a DNS IP address that you want to delete. Select the DNS IP Address checkbox when you want to delete all IP addresses. You must select Manually configure DNS list before you can add or delete DNS IP addresses.
System Settings Administrator page The Administrator page allows you to configure the administrator password and how the system communicates with the administrator.
Table 7 Administrator page settings
User Name The default administrator name is sysadmin. The sysadmin user cannot be renamed or deleted. You can create other users (including other administrators) and change the administrator password by selecting Administration > Access > Local Users. Password Type the password for sysadmin. Verify Password Retype the password for sysadmin. Admin Email Specify the email address to which DD System Manager sends alert and autosupport email messages. Send Alert Notification Emails Check to configure DD System Manager to send alert to this address notifications to sysadmin as alert events occur. Send Daily Alert Summary Check to configure DD System Manager to send alert summaries Emails to this address to sysadmin at the end of each day. Send Autosupport Emails to this Check to configure DD System Manager to send the sysadmin address user autosupport emails, which are daily reports that document system activity and status.
System Settings Administrator page 37 Getting Started
System Settings Email/Location page The Email/Location page allows you to configure the mail server name, control what system information is sent to Data Domain, and specify a location name to identify your system.
Table 8 Email/Location page settings
Mail Server Specify the name of the mail server that manages emails to and from the system. Send Alert Notification Emails Check to configure DD System Manager to send alert to Data Domain notification emails to Data Domain. Send Vendor Support Check to configure DD System Manager to send vendor support Notification Emails to Data notification emails to Data Domain. Domain
Location Use this optional attribute as needed to record the location of your system. If you specify a location, this information is stored as the SNMP system location.
CIFS Protocol Authentication page The Authentication page enables you to configure Active Directory, Workgroup and NIS authentication information for your system. To configure these settings outside of the configuration wizard, select Administration > Access > Authentication.
Table 9 Authentication page settings
Active Directory/Kerberos Expand this panel to enable, disable, and configure Active Authentication Directory Kerberos authentication. Workgroup Authentication Expand this panel to configure Workgroup authentication. NIS Authentication Expand this panel to enable, disable, and configure NIS authentication, servers, and groups.
CIFS Protocol Share page The Share page enables you to configure a CIFS protocol share name and a directory path for your system. To configure these settings outside of the configuration wizard, select Protocols > CIFS > Shares > Create.
38 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
Table 10 Share page settings
Share Name Enter a share name for the system. Directory Path Enter a directory path for the system. Add (+) button Click + to enter a system client. Delete (X) button Click X to delete a selected client.
NFS Protocol Export page The Export page enables you to configure an NFS protocol export directory path and network clients. To configure these settings outside of the configuration wizard, select Protocols > NFS > Create.
Table 11 Export page settings
Directory Path Enter a pathname for the export. Add (+) button Click + to enter a system client. Delete (X) button Click X to delete a selected client.
DD Boost Protocol Storage Unit page The Storage Unit page allows you to configure DD Boost storage units. To configure these settings outside of the configuration wizard, select Protocols > DD Boost > Storage Units > + (plus sign) to add a storage unit, the pencil to modify a storage unit, or X to delete a storage unit.
Table 12 Storage Unit page settings
Storage Unit The name of your DD Boost Storage Unit. You may optionally change this name. User For the default DD Boost user, either select an existing user, or select Create a new Local User, and enter their User name, Password, and Management Role. This role can be one of the following:
l Admin role: Lets you configure and monitor the entire Data Domain system.
l User role: Lets you monitor Data Domain systems and change your own password.
l Security role: In addition to user role privileges, lets you set up security-officer configurations and manage other security-officer operators.
NFS Protocol Export page 39 Getting Started
Table 12 Storage Unit page settings (continued)
l Backup-operator role: In addition to user role privileges, lets you create snapshots, import and export tapes to a VTL, and move tapes within a VTL.
l None role: Intended only for EMC DD Boost authentication, so you cannot monitor or configure a Data Domain system.
DD Boost Protocol Fibre Channel page The Fibre Channel page allows you to configure DD Boost Access Groups over Fibre Channel. To configure these settings outside of the configuration wizard, select Protocols > DD Boost > Fibre Channel > + (plus sign) to add an access group, the pencil to modify an access group, or X to delete an access group.
Table 13 Fibre Channel page settings
Configure DD Boost over Fibre Select the checkbox if you want to configure DD Boost over Channel Fibre Channel. Group Name (1-128 Chars) Create an Access Group. Enter a unique name. Duplicate access groups are not supported. Initiators Select one or more initiators. Optionally, replace the initiator name by entering a new one. An initiator is a backup client that connects to the system to read and write data using the FC (Fibre Channel) protocol. A specific initiator can support DD Boost over FC or VTL, but not both. Devices The devices to be used are listed. They are available on all endpoints. An endpoint is the logical target on the Data Domain system to which the initiator connects.
VTL Protocol Library page The Library page allows you to configure VTL protocol settings for a library. To configure these settings outside of the configuration wizard, select PROTOCOLS > VTL > Virtual Tape Libraries > VTL Service > Libraries > More Tasks > Library > Create
Table 14 Library page settings
Library Name Enter a name of from 1 to 32 alphanumeric characters. Number of Drives Number of supported tape drives. Drive Model Select the desired model from the drop-down list:
40 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
Table 14 Library page settings (continued)
l IBM-LTO-5 (default)
Number of Slots Enter the number of slots per library:
l Up to 32,000 slots per library
l Up to 64,000 slots per system
l This should be equal to, or greater than, the number of drives.
Number of CAPs (Optional) Enter the number of cartridge access ports (CAPs):
l Up to 100 CAPs per library
l Up to 1000 CAPs per system
Changer Model Name Select the desired model from the drop-down list:
Starting Barcode Enter the desired barcode for the first tape, in the format A990000LA. Tape Capacity (Optional) Enter the tape capacity. If not specified, the capacity is derived from the last character of the barcode.
VTL Protocol Access Group page The Access Group page allows you to configure VTL protocol settings for an access group. To configure these settings outside of the configuration wizard, select PROTOCOLS > VTL > Access Groups > Groups > More Tasks > Group > Create.
Table 15 Access Group page settings
Group Name Enter a unique name of from 1 - 128 characters. Duplicate access groups are not supported. Initiators Select one or more initiators. Optionally, replace the initiator name by entering a new one. An initiator is a backup client that connects to a system
VTL Protocol Access Group page 41 Getting Started
Table 15 Access Group page settings (continued)
to read and write data using the Fibre Channel (FC) protocol. A specific initiator can support EMC DD Boost over FC or VTL, but not both. Devices The devices (drives and changer) to be used are listed. These are available on all endpoints. An endpoint is the logical target on the Data Domain system to which the initiator connects.
Data Domain Command Line Interface The Command Line Interface (CLI) is a text driven interface that can be used instead of or in addition to DD System Manager. Most management tasks can be performed in DD System Manager or with the CLI. In some cases, the CLI offers configuration options and reports that are not yet supported in DD System Manager. Any Data Domain system command that accepts a list, such as a list of IP addresses, accepts entries separated by commas, by spaces, or both. The Tab key can be used to do the following. l Complete a command entry when that entry is unique. Tab completion is supported for all keywords. For example, entering syst Tab shTab st Tab displays the command system show stats. l Show the next available option, if you do not enter any characters before pressing the Tab key. l Show partial matched tokens or complete a unique entry, if you enter characters before pressing the Tab key. The EMC Data Domain Operating System Command Reference Guide provides information for each of the CLI commands. Online help is available and provides the complete syntax for each command.
Logging into the CLI You can access the CLI using a direct connection to the system or using an Ethernet connection through SSH or Telenet. Before you begin To use the CLI, you must establish a local or remote connection to the system using one of the following methods. l If you are connecting through a serial console port on the system, connect a terminal console to the port and use the communication settings: 9600 baud, 8 data bits, no parity, and 1 stop bit. l If the system provides keyboard and monitor ports, connect a keyboard and monitor to those ports. l If you are connecting through Ethernet, connect a computer with SSH or Telnet client software to an Ethernet network that can communicate with the system. Procedure 1. If you are using an SSH or Telnet connection to access the CLI, start the SSH or Telnet client and specify the IP address or host name of the system.
42 EMC Data Domain Operating System 5.7 Administration Guide Getting Started
For information on initiating the connection, see the documentation for the client software. The system prompts you for your username. 2. When prompted, enter your system username. 3. When prompted, enter your system password.
The following example shows SSH login to a system named mysystem using SSH client software.
# ssh -l sysadmin mysystem.mydomain.com Data Domain OS 5.6.0.0-19899 Password:
CLI online help guidelines The CLI displays two types of help, syntax-only help and command-description help that includes the command syntax. Both types of help offer features that allow you reduce the time it takes to find the information you need. The following guidelines describe how to use syntax-only help.
l To list the top-level CLI commands, enter a question mark (?), or type the command help at the prompt.
l To list all forms of a top-level command, enter the command with no options at the prompt or enter command ?. l To list all commands that use a specific keyword, enter help keyword or ? keyword. For example, ? password displays all Data Domain system commands that use the password argument. The following guidelines describe how to use command-description help.
l To list the top-level CLI commands, enter a question mark (?), or type the command help at the prompt. l To list all forms of a top-level command with an introduction, enter help command or ? command. l The end of each help description is marked END. Press Enter to return to the CLI prompt. l When the complete help description does not fit in the display, the colon prompt (:) appears at the bottom of the display. The following guidelines describe what you can do when this prompt appears.
n To move through the help display, use the up and down arrow keys. n To quit the current help display and return to the CLI prompt, press q.
n To display help for navigating the help display, press h.
n To search for text in the help display, enter a slash character (/) followed by a pattern to use as search criteria and press Enter. Matches are highlighted.
CLI online help guidelines 43 Getting Started
44 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 3 Managing Data Domain Systems
This chapter includes:
l System management overview. 46 l Rebooting a system. 47 l Powering a system on or off . 47 l System upgrade management. 48 l Managing system licenses. 50 l System storage management. 51 l Network connection management. 56 l System passphrase management. 77 l System access management. 79 l Configuring mail server settings. 101 l Managing time and date settings. 101 l Managing system properties. 102 l SNMP management. 103 l Autosupport report management. 111 l Support bundle management. 113 l Alert notification management. 114 l EMC Support delivery management. 120 l Log file management. 122 l Remote system power management with IPMI. 127
Managing Data Domain Systems 45 Managing Data Domain Systems
System management overview DD System Manager allows you to manage the system on which DD System Manager is installed.
l To support replication, DD System Manager supports the addition of systems running the previous two versions, the current version and the next two versions as they become available. For Release 5.7.1, DD System Manager supports the addition of systems for replication for DD OS Version 5.5 to 5.7 plus the next two releases.
When processing a heavy load, a system might be less responsive than normal. In this case, management commands issued from either DD System Manager or the CLI might take longer to complete. When the duration exceeds allowed limits, a timeout error is returned, even if the operation completed.
The following table recommends the maximum number of user sessions supported by DD System Manager:
Table 16 Maximum number of users supported by DD System Manager
System Model Maximum Active Users Maximum Logged In Users
4 GB modelsa 5 10 8 GB modelsb 10 15 16 GB and greater modelsc 10 20
a. Includes DD140 and DD2200 (4TB) b. Includes DD610 and DD630 c. Includes DD670, DD860, DD890, DD990, DD2200 (>7.5TB), DD4200, DD4500, DD7200, and DD9500
Initial HA system set-up (DD9500 only) cannot be done from the DD System Manager, but the status of an already-configured HA system can be viewed from the DDSM.
HA system management overview The HA relationship between the two nodes, one active and one standby, is setup through DDSH CLIs. Initial set-up can be run on either of the two nodes but only one at a time. It is a precondition of HA that the system interconnect and identical hardware is setup on both nodes first.
Both DDRs are required to have identical hardware which will be validated during setup and system boot-up.
If set-up is from a fresh install of systems, the ha create command needs to be run on the node with the license installed. If setup is from an existing system and a new fresh install system (upgrade), then it should be run from the existing system.
46 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
HA system planned maintenance In order to reduce planned maintenance downtime, system upgrade is included in the HA architecture. A rolling system upgrade may use a manual HA failover to move the services away from the node that must be upgraded. That node may then be shut down and upgraded. The newly upgraded node is then rebooted to resume redundant operations as the standby node. The upgrade is then repeated for the active node. It is shutdown, allowing the newly upgraded standby node to take over. Once the upgrade is performed on the second node, it is rebooted and assumes the role of the standby node. All system upgrade operations that may require data conversion may not start until both systems are upgraded to the same level and HA state is fully restored.
Rebooting a system Reboot a system when a configuration change, such as changing the time zone, requires that you reboot the system. Procedure
1. Select Maintenance > System > More Tasks > Reboot System. 2. Click OK to confirm.
Powering a system on or off When powering a system off and on, it is important that you follow the proper procedure to preserve the file system and configuration integrity. Procedure 1. Power on any expansion shelves before powering on the Data Domain controller. Wait approximately three minutes after all expansion shelves are turned on.
A controller is the chassis and any internal storage. A Data Domain system refers to the controller and any optional external storage.
2. Plug in the power cord for your controller, and if there is a power button on the controller, press the power button (as shown in the Installation and Setup Guide for your Data Domain system). 3. To shut down power to a Data Domain system, use the system poweroff CLI command. This command automatically performs an orderly shut down of DD OS processes and is available to administrative users only.
Do not use the chassis power switch to power off the system. Doing so prevents remote power control using IPMI. Use the system poweroff command instead. The system poweroff command shuts down the system and turns off the power.
HA system planned maintenance 47 Managing Data Domain Systems
The IMPI Remote System Power Down feature does not perform an orderly shutdown of the DD OS. Use this feature only if the system poweroff command is unsuccessful.
System upgrade management To upgrade a DD OS system, you must verify that there is sufficient room for the new software on the target system, transfer the software to the system to be upgraded, and then start the upgrade. For HA systems, use the floating IP address to access DD System Manager to perform software upgrades.
Viewing upgrade packages on the system DD System Manager allows you to view and manage up to five upgrade packages on a system. Before you can upgrade a system, you must download an upgrade package from the EMC Online Support site to a local computer, and then upload it to the target system. Procedure
1. Select Maintenance > System. Results For every package stored on the system, DD System Manager displays the filename, file size, and last modified date in the list titled: Upgrade Packages Available on Data Domain System.
Obtaining and verifying upgrade packages You can use DD System Manager to locate upgrade package files on the Data Domain Support Web site and upload copies of those files to a system.
You can use FTP or NFS to copy an upgrade package to a system. DD System Manager is limited to managing 5 system upgrade packages, but there are no restrictions, other than space limitations, when you manage the files directly in the /ddvar/releases directory. FTP is disabled by default. To use NFS, /ddvar needs to be exported and mounted from an external host).
1. Select Maintenance > System. 2. To obtain an upgrade package, click the EMC Online Support link, click Downloads, and use the search function to locate the package recommended for your system by Support personnel. Save the upgrade package to the local computer. 3. Verify that there are no more than four packages listed in the Upgrade Packages Available on Data Domain System list. DD System Manager can manage up to five upgrade packages. If five packages appear in the list, remove at least one package before uploading the new package. 4. Click Upload Upgrade Package to initiate the transfer of the upgrade package to the system.
48 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
5. In the Upload Upgrade Package dialog, click Browse to open the Choose File to Upload dialog. Navigate to the folder with the downloaded file, select the file, and click Open. 6. Click OK. An upload progress dialog appears. Upon successful completion of the upload, the download file (with a .rpm extension) appears in the list titled: Upgrade Packages Available on Data Domain System.
7. To verify the upgrade package integrity, click View Checksum and compare the calculated checksum displayed in the dialog to the authoritative checksum on the EMC Online Support site. 8. To manually initiate an upgrade precheck, select an upgrade package and click Upgrade Precheck.
Upgrading a Data Domain system When an upgrade package file is present on a system, you can use DD System Manager to perform an upgrade using that upgrade package. Before you begin Read the DD OS Release Notes for the complete upgrade instructions and coverage of all the issues that can impact the upgrade. The procedure that follows describes how to initiate an upgrade using DD System Manager.
Upgrade package files use the .rpm file extension. This topic assumes that you are updating only DD OS. If you make hardware changes, such as adding, swapping, or moving interface cards, you must update the DD OS configuration to correspond with the hardware changes.
Procedure 1. Log into the system where the upgrade is to be performed.
For most releases, upgrades are permitted from up to two prior major release versions. For Release 5.7, upgrades are permitted from Releases 5.5 and 5.6.
As recommended in the Release Notes, reboot the Data Domain system before upgrading to verify that the hardware is in a clean state. If any issues are discovered during the reboot, resolve those issues before starting the upgrade.
2. Select Data Management > File System, and verify that the file system is enabled and running. 3. Select Maintenance > System. 4. From the Upgrade Packages Available on Data Domain System list, select the package to use for the upgrade.
Upgrading a Data Domain system 49 Managing Data Domain Systems
You must select an upgrade package for a newer version of DD OS. DD OS does not support downgrades to previous versions.
5. Click Perform System Upgrade. The System Upgrade dialog appears and displays information about the upgrade and a list of users who are currently logged in to the system to be upgraded.
6. Verify the version of the upgrade package, and click OK to continue with the upgrade. The System Upgrade dialog displays the upgrade status and the time remaining. When upgrading the system, you must wait for the upgrade to complete before using DD System Manager to manage the system. If the system restarts, the upgrade might continue after the restart, and DD System Manager displays the upgrade status after login. EMC recommends that you keep the System Upgrade progress dialog open until the upgrade completes or the system powers off. When upgrading DD OS Release 5.5 or later to a newer version, and if the system upgrade does not require a power off, a Login link appears when the upgrade is complete.
To view the status of an upgrade using the CLI, enter the system upgrade status command. Log messages for the upgrade are stored in /ddvar/log/ debug/platform/upgrade-error.log and /ddvar/log/debug/ platform/upgrade-info.log.
7. If the system powers down, you must remove AC power from the system to clear the prior configuration. Unplug all of the power cables for 30 seconds and then plug them back in. The system powers on and reboots. 8. If the system does not automatically power on and there is a power button on the front panel, press the button.
Removing an upgrade package A maximum of five upgrade packages can be uploaded to a system with DD System Manager. If the system you are upgrading contains five upgrade packages, you must remove at least one package before you can upgrade the system. Procedure
1. Select Maintenance > System. 2. From the list titled Upgrade Packages Available on Data Domain System, select the package to remove. One package can be removed at a time. 3. Click Remove Upgrade Package.
Managing system licenses Add system licenses using purchased license keys to enable Data Domain system features. Remove a system license when you are decommissioning a system. Procedure
1. Select Administration > Licenses.
50 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
The Feature Licenses list displays the installed feature licenses and names. The Shelf Capacity Licenses list displays licenses that enable the use of additional storage shelves.
2. To add a license, complete the following tasks. a. Click Add Licenses. b. In the Add Licenses Key dialog box, type or paste one or more license keys in the License Key box. Type each key on its own line, or separate each key by a space or comma (DD System Manager automatically places each key on a new line). c. Click Add. The added licenses display in the Added license list. Any errors are listed in the error license list. Select a license with an error to edit it. Select Retry Failed License(s) to retry the key. Otherwise, click Done to ignore the errors and return to the Feature Licenses list.
3. To delete a license, complete the following tasks. a. In the Feature Licenses list, select one or more licenses to remove and click Delete Selected Licenses. b. In the Warning dialog box verify the license(s) to delete and click OK. The licenses are removed from the license list.
HA system license management HA is a licensed feature, and the system licensing key is registered by following the same steps to add any other license to the DD system. A system will be configured as Active-Standby, where one node is designated "standby." Only one set of licenses will be required rather than needing individual licenses for each node. During failover, the licenses on one node will failover to the other node.
System storage management System storage management features allow you to view the status and configuration of your storage space, flash a disk LED to facilitate disk identification, and change the storage configuration.
All storage connected or used by the two-node Active-Standby HA system can be viewed as a single system.
Viewing system storage information The storage status area shows the current status of the storage, such as Operational or Non-Operational, and the storage migration status. Below the Status area are tabs that organize how the storage inventory is presented. Procedure
1. To display the storage status, select Hardware > Storage. 2. If an alerts link appears after the storage status, click the link to view the storage alerts.
HA system license management 51 Managing Data Domain Systems
3. If the Storage Migration Status is Not licensed, you can click Add License to add the license for this feature.
Overview tab The Overview tab displays information for all disks in the Data Domain system organized by type. The categories that display are dependent on the type of storage configuration in use. The Overview tab lists the discovered storage in one or more of the following sections.
l Active Tier Disks in the Active Tier are currently marked as usable by the Data Domain file system. Disks are listed in two tables, Disks in Use and Disks Not in Use. l Retention Tier If the optional EMC Data Domain Extended Retention (formerly DD Archiver) license is installed, this section shows the disks that are configured for Extended Retention storage. Disks are listed in two tables, Disks in Use and Disks Not in Use. For more information, see the EMC Data Domain Extended Retention Administration Guide. l Usable Disks and Enclosures For systems with optional enclosures, this section shows the disks and enclosures that can be added to the system. l Failed/Foreign/Absent Disks (Excluding Systems Disks) Shows the disks that are in a failed state; these cannot be added to the system Active or Retention tiers. l Systems Disks Shows the disks where the DD OS resides when the Data Domain controller does not contain data storage disks. Each section heading displays a summary of the storage configured for that section. The summary shows tallies for the total number of disks, disks in use, spare disks, reconstructing spare disks, available disks, and known disks. Click a section plus (+) button to display detailed information, or click the minus (-) button to hide the detailed information.
Table 17 Disks In Use column label descriptions
Disk Group The name of the disk group that was created by the file system (for example, dg1). Status The status of the disk (for example Normal, Warning). Disks Reconstructing The disks that are undergoing reconstruction, by disk ID (for example, 1.11). Total Disks The total number of usable disks (for example, 14). Disks The disk IDs of the usable disks (for example, 2.1-2.14).
Table 18 Disks Not In Use column label descriptions
Disk The disk identifier, which can be any of the following.
l The enclosure and disk number (in the form Enclosure Slot)
52 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 18 Disks Not In Use column label descriptions (continued)
l A device number for a logical device such as those used by VTL and vDisk
Status The status of the disk, for example In Use, Available, Spare. Size The data storage capacity of the disk when used in a Data Domain system.a Manufacturer/Model The manufacturer’s model designation. The display may include a model ID or RAID type or other information depending on the vendor string sent by the storage array. Firmware The firmware level used by the third-party physical-disk storage controller.
Serial Number The manufacturer’s serial number for the disk.
a. The Data Domain convention for computing disk space defines one gibibyte as 230 bytes, giving a different disk capacity than the manufacturer’s rating.
Status tab The Status tab displays a summary status table showing the state of all system disks and another table that provides additional information on reconstructing disks.
Table 19 Disks Status table column label descriptions
Total The total number of inventoried disks in the Data Domain system. In Use The number of disks currently in use by the file system. Spare The number of spare disks (available to replace failed disks). Spare (reconstructing) The number of disks that are in the process of data reconstruction (spare disks replacing failed disks). Available The number of disks that are available for allocation to an Active or Extended Retention storage tier. Known The number of known unallocated disks. Unknown The number of unknown unallocated disks. Failed The number of failed disks. Foreign The number of foreign disks. Absent The number of absent disks. Migrating The number of disks serving as the source of a storage migration. Destination The number of disks serving as the destination of a storage migration. Not Installed The number of empty disk slots that the system can detect.
Viewing system storage information 53 Managing Data Domain Systems
The following table describes the entries in the Reconstructing table.
Table 20 Reconstruction table column label descriptions
Disk Identifies disks that are being reconstructed. Disk labels are of the format enclosure.disk. Enclosure 1 is the Data Domain system, and external shelves start numbering with enclosure 2. For example, the label 3.4 is the fourth disk in the second shelf. Disk Group Shows the RAID group (dg#) for the reconstructing disk. Tier The name of the tier where the failed disk is being reconstructed. Time Remaining The amount of time before the reconstruction is complete. Percentage Complete The percentage of reconstruction that is complete.
When a spare disk is available, the Data Domain file system automatically replaces a failed disk with a spare and begins the reconstruction process to integrate the spare into the RAID disk group. The disk use displays Spare and the status becomes Reconstructing. Reconstruction is performed on one disk at a time.
Disks tab The Disks tab displays information on each of the system disks. You can filter the disks viewed to display all disks, disks in a specific tier, or disks in a specific group.
Table 21 Disks Status table column label descriptions
Disk The disk identifier, which can be:
l The enclosure and disk number (in the form Enclosure.Slot).
l A device number for a logical device such as those used by VTL and vDisk..
Status The status of the disk, which can be one of the following.
l Absent. No disk is installed in the indicated location.
l Available. An available disk is allocated to the active or retention tier, but it is not currently in use.
l Copy Recovery. The disk has a high error rate but is not failed. RAID is currently copying the contents onto a spare drive and will fail the drive once the copy reconstruction is complete.
l Destination. The disk is in use as the destination for storage migration.
l Error. The disk has a high error rate but is not failed. The disk is in the queue for copy reconstruction. The state will change to Copy Recovery when copy reconstruction begins.
l Foreign. The disk has been assigned to a tier, but the disk data indicates the disk may be owned by another system.
54 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 21 Disks Status table column label descriptions (continued)
l In-Use. The disk is being used for backup data storage.
l Known. The disk is a supported disk that is ready for allocation.
l Migrating. The disk is in use as the source for storage migration.
l Powered Off. The disk power has been removed by EMC Support.
l Reconstruction. The disk is reconstructing in response to a disk fail command or by direction from RAID/SSM.
l Spare. The disk is available for use as a spare.
l System. System disks store DD OS and system data. No backup data is stored on system disks.
l Unknown. An unknown disk is not allocated to the active or retention tier. It might have been failed administratively or by the RAID system.
Manufacturer/Model The manufacturer’s model designation. The display may include a model ID or RAID type or other information depending on the vendor string sent by the storage array. Firmware The firmware level used by the third-party physical-disk storage controller. Serial Number The manufacturer’s serial number for the disk.
Physically locating an enclosure If you have trouble determining which physical enclosure corresponds to an enclosure displayed in DD System Manager, you can use the CLI beacon feature to flash the enclosure IDENT LEDs and all the disk LEDs that indicate normal operation. Procedure 1. Establish a CLI session with the system. 2. Type enclosure beacon enclosure.
3. Press Ctrl-C to stop the LED flashing.
Physically locating a disk If you have trouble determining which physical disk corresponds to a disk displayed in DD System Manager, you can use the beacon feature to flash an LED on the physical disk. Procedure
1. Select Hardware > Storage > Disks. 2. Select a disk from the Disks table and click Beacon.
Physically locating an enclosure 55 Managing Data Domain Systems
You can select one disk at a time.
The Beaconing Disk dialog box appears, and the LED light on the disk begins flashing.
3. Click Stop to stop the LED beaconing.
Configuring storage Storage configuration features allow you add and remove storage expansion enclosures from the active and retention tiers. Storage in an expansion enclosure (which is sometimes called an expansion shelf) is not available for use until it is added to a tier.
Additional storage requires the appropriate license or licenses and sufficient memory to support the new storage capacity. Error messages display if more licenses or memory is needed.
Procedure 1. Select Hardware > Storage > Overview > Configure Storage. 2. In the Configure Storage dialog, select the storage to be added from the Available Storage list. 3. In the Configure list, select either Active Tier or Retention Tier. The maximum amount of storage that can be added to the active tier depends on the DD controller used.
The licensed capacity bar shows the portion of licensed capacity (used and remaining) for the installed enclosures.
4. Select the checkbox for the Shelf to be added. 5. Click the Add to Tier button. 6. Click OK to add the storage.
To remove an added shelf, select it in the Tier Configuration list, click Remove from Configuration, and click OK.
Network connection management Network connection management features allow you view and configure network interfaces, general network settings, and network routes.
HA system network connection management The HA system relies on two different types of IP addresses, fixed and floating. Each type has specific behaviors and limitations. On an HA system, Fixed IP addresses:
56 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
l Are used for node management via the CLI l Are attached ("fixed") to the node l Can be static or DHCP, IPv6 SLAAC l Configuration is done on the specific node with the optional type fixed argument
All filesystem access should be through a floating IP.
Floating IP addresses only exist in the two-node HA system; during failover, the IP address "float" to the new active node and are:
l Only configured on the active node l Used for filesystem access and most configuration l Can only be static l Configuration requires the type floating argument
Network interface management Network interface management features allow you to manage the physical interfaces that connect the system to a network and create logical interfaces to support link aggregation, load balancing, and link or node failover.
Viewing interface information The Interfaces tab allows you to manage physical and virtual interfaces, VLANs, DHCP, DDNS, and IP addresses and aliases. Consider the following guidelines when managing IPv6 interfaces.
l The command-line interface (CLI) supports IPv6 for basic Data Domain network and replication commands, but not for backup and Extended Retention (archive) commands. CLI commands manage the IPv6 addresses. You can view IPv6 addresses using the DD System Manager, but you cannot manage IPv6 with the DD System Manager. l Collection, directory, and MTree replication are supported over IPv6 networks, which allows you to take advantage of the IPv6 address space. Simultaneous replication over IPv6 and IPv4 networks is also supported, as is Managed File Replication using DD Boost. l There are some restrictions for interfaces with IPv6 addresses. For example, the minimum MTU is 1280. If you try to set the MTU lower than 1280 on an interface with an IPv6 address, an error message appears and the interface is removed from service. An IPv6 address can affect an interface even though it is on a VLAN attached to the interface and not directly on the interface. Procedure
1. Select Hardware > Ethernet > Interfaces. The following table describes the information on the Interfaces tab.
Table 22 Interface tab label descriptions
Interface The name of each interface associated with the selected system.
Network interface management 57 Managing Data Domain Systems
Table 22 Interface tab label descriptions (continued)
Enabled Whether the interface is enabled.
l Select Yes to enable the interface and connect it to the network. l Select No to disable the interface and disconnect it from the network.
DHCP Indicates if the interface is configured manually (no), by a DHCP (Dynamic Host Configuration Protocol) IPv4 server (v4), or by a DHCP IPv6 server (v6).
IP Address IP address associated with the interface. The address used by the network to identify the interface. If the interface is configured through DHCP, an asterisk appears after this value.
Netmask Netmask associated with the interface. Uses the standard IP network mask format. If the interface is configured through DHCP, an asterisk appears after this value. Link Whether the Ethernet connection is active (Yes/No). Address Type On an HA system, the Address Type indicates Fixed, Floating, or Interconnect. Additional Info Additional settings for the interface. For example, the bonding mode. IPMI interfaces Displays Yes or No and indicates if IPMI health monitoring and power configured management is configured for the interface.
2. To filter the interface list by interface name, enter a value in the Interface Name field and click Update. Filters support wildcards, such as eth*, veth*, or eth0*
3. To filter the interface list by interface type, select a value from the Interface Type menu and click Update. On an HA system, there is a filter dropdown to filter by IP Address Type (Fixed, Floating, or Interconnect). 4. To return the interfaces table to the default listing, click Reset. 5. Select an interface in the table to populate the Interface Details area.
Table 23 Interface Details label descriptions
Auto-generated Displays the automatically generated IPv6 addresses for the selected Addresses interface. Auto Negotiate When this feature displays Enabled, the interface automatically negotiates Speed and Duplex settings. When this feature displays Disabled, then Speed and Duplex values must be set manually.
Cable Shows whether the interface is Copper or Fiber.
58 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 23 Interface Details label descriptions (continued)
Some interfaces must be up before the cable status is valid.
Duplex Used in conjunction with the Speed value to set the data transfer protocol. Options are Unknown, Full, Half. Hardware Address The MAC address of the selected interface. For example, 00:02:b3:b0:8a:d2. Interface Name Name of the selected interface. Latent Fault Detection The LFD field has a View Configuration link, displaying a pop- (LFD) - HA systems only up that lists LFD addresses and interfaces.
Maximum Transfer Unit MTU value assigned to the interface. (MTU) Speed Used in conjunction with the Duplex value to set the rate of data transfer. Options are Unknown, 10 Mb/s, 100 Mb/s, 1000 Mb/s, 10 Gb/s.
Auto-negotiated interfaces must be set up before speed, duplex, and supported speed are visible.
Supported Speeds Lists all of the speeds that the interface can use.
6. To view IPMI interface configuration and management options, click View IPMI Interfaces. This link displays the Maintenance > IPMI information.
Physical interface names and limitations The format of physical interface names varies on different Data Domain systems and option cards, and limitations apply to some interfaces. l For most systems the physical interface name format is ethxy, where x is the slot number for an on-board port or an option card and y is an alphanumeric string. For example, eth0a. l For most on-board NIC vertical interfaces, the top interface is named eth0a and the bottom interface is eth0b. l For most on-board NIC horizontal interfaces, the left interface as viewed from the rear, is named eth0a and the right is named eth0b. l DD990 systems provide four on-board interfaces: two on the top and two on the bottom. The top-left interface is eth0a, the top-right is eth0b, the bottom-left is eth0c, and the bottom-right is eth0d. l DD2200 systems provide four on-board 1G Base-T NIC ports: ethMa (top left), ethMb (top right), ethMc (bottom left), and ethMd (bottom right).
Network interface management 59 Managing Data Domain Systems
l DD2500 systems provide six on-board interfaces. The four on-board 1G Base-T NIC ports are ethMa (top left), ethMb (top right), ethMc (bottom left), and ethMd (bottom right). The two on-board 10G Base-T NIC ports are ethMe (top) and ethMf (bottom). l DD4200, DD4500, and DD7200 systems provide one on-board Ethernet port, which is ethMa. l For systems ranging between DD140 and DD990, the physical interface names for I/O modules start at the top of the module or at the left side. The first interface is ethxa, the next is ethxb, the next is ethxc, and so forth. l The port numbers on the horizontal DD2500 I/O modules are labeled in sequence from the end opposite the module handle (left side). The first port is labeled 0 and corresponds to physical interface name ethxa, the next is 1/ethxb, the next is 2/ ethxc, and so forth. l The port numbers on the vertical DD4200, DD4500, and DD7200 I/O modules are labeled in sequence from the end opposite the module handle (bottom). The first port is labeled 0 and corresponds to physical interface name ethxa, the next is 1/ethxb, the next is 2/ethxc, and so forth.
General interface configuration guidelines Review the general interface configuration guidelines before configuring system interfaces.
l When supporting both backup and replication traffic, EMC recommends using different interfaces for each traffic type so that neither traffic type impacts the other. l When replication traffic is expected to be less than 1 Gb/s, EMC recommends against using 10 GbE interfaces for replication traffic because 10 GbE interfaces are optimized for faster traffic. l On DD4200, DD4500, and DD7200 systems that use IPMI, EMC recommends that interface ethMa be reserved for IPMI traffic and system management traffic (using protocols such as HTTP, Telnet, and SSH). Backup data traffic should be directed to other interfaces.
Configuring physical interfaces You must configure at least on physical interface before the system can connect to a network. Procedure
1. Select Hardware > Ethernet > Interfaces. 2. Select an interface to configure.
DD140, DD160, DD610, DD620, and DD630 systems do not support IPv6 on interface eth0a (eth0 on systems that use legacy port names) or on any VLANs created on that interface.
3. Click Configure. 4. In the Configure Interface dialog, determine how the interface IP address is to be set:
60 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
On an HA system, the Configure Interface dialog has a field for whether or not to designate the Floating IP (Yes/No). Selecting Yes the Manually Configure IP Address radio button is auto-selected; Floating IP interfaces can only be manually configured.
l Use DHCP to assign the IP address—in the IP Settings area, select Obtain IP Address using DHCP and select either DHCPv4 for IPv4 access or DHCPv6 for IPv6 access. Setting a physical interface to use DHCP automatically enables the interface.
If you choose to obtain the network settings through DHCP, you can manually configure the hostname at Hardware > Ethernet > Settings or with the net set hostname command. You must manually configure the host name when using DHCP over IPv6.
l Specify IP Settings manually—in the IP Settings area, select Manually configure IP Address. The IP Address and Netmask fields become active.
5. If you chose to manually enter the IP address, enter an IPv4 or IPv6 address. If you entered an IPv4 address, enter a netmask address.
You can assign just one IP address to an interface with this procedure. If you assign another IP address, the new address replaces the old address. To attach an additional IP address to an interface, create an IP alias.
6. Specify Speed/Duplex settings. The combination of speed and duplex settings define the rate of data transfer through the interface. Select one of these options: l Autonegotiate Speed/Duplex — Select this option to allow the network interface card to autonegotiate the line speed and duplex setting for an interface. Autonegotiation is not supported on the following DD2500, DD4200, DD4500, and DD7200 I/O modules: n Dual Port 10GbE SR Optical with LC connectors (using SFPs) n Dual Port 10GbE Direct Attach Copper (SFP+ cables) n Quad port 2 port 1GbE Copper (RJ45) /2 port 1GbE SR Optical l Manually configure Speed/Duplex — Select this option to manually set an interface data transfer rate. Select the speed and duplex from the menus. n Duplex options are half-duplex, full-duplex, and unknown. n Speed options listed are limited to the capabilities of the hardware device. Options are 10 Mb, 100 Mb, 1000 Mb (1 Gb), 10 Gb, and unknown. The 10G Base-T hardware supports only the 100 Mb, 1000 Mb and 10 Gb settings. n Half-duplex is only available for 10 Mb and 100 Mb speeds. n 1000 Mb and 10 Gb line speeds require full-duplex. n On DD2500, DD4200, DD4500, and DD7200 10GbE I/O modules, copper interfaces support only the 10 Gb speed setting.
Network interface management 61 Managing Data Domain Systems
n The default setting for 10G Base-T interfaces is Autonegotiate Speed/Duplex. If you manually set the speed to 1000 Mb or 10 Gb, you must set the Duplex setting to Full.
7. Specify the MTU (Maximum Transfer Unit) size for the physical (Ethernet) interface. Do the following:
l Click the Default button to return the setting to the default value. l Ensure that all of your network components support the size set with this option.
8. Optionally, select Dynamic DNS Registration. Dynamic DNS (DDNS) is a protocol that registers local IP addresses on a Domain Name System (DNS) server. In this release, DD System Manager supports Windows mode DDNS. To use UNIX mode DDNS, use the net ddns CLI command. The DDNS must be registered to enable this option.
This option disables DHCP for this interface.
9. Click Next. The Configure Interface Settings summary page appears. The values listed reflect the new system and interface state, which are applied after you click Finish.
10. Click Finish and OK. MTU size values The MTU size must be set properly to optimize the performance of a network connection. An incorrect MTU size can negatively affect interface performance. Supported values for setting the maximum Transfer Unit (MTU) size for the physical (Ethernet) interface range from 350 to 9000. For 100 Base-T and gigabit networks, 1500 is the standard default.
Although 9000 is the maximum size, to be backwardly compatible, the DD OS accepts up to 9014, but sets the size to 9000 when it is greater than 9000 and less than or equal to 9014.
The minimum MTU for IPv6 interfaces is 1280. The interface fails if you try to set the MTU lower than 1280.
Moving a static IP address A specific static IP address must be assigned to only one interface on a system. A static IP address must be properly removed from an interface before it is configured on another interface. Procedure 1. If the interface that hosts the static IP address is part of a DD Boost interface group, remove the interface from that group.
62 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
2. Select Hardware > Ethernet > Interfaces. 3. Remove the static IP address that you want to move. a. Select the interface that is currently using the IP address you want to move. b. In the Enabled column, select No to disable the interface. c. Click Configure. d. Set the IP Address to 0.
Set the IP address to 0 when there is no other IP address to assign to the interface. The same IP address must not be assigned to multiple interfaces.
e. Click Next, and click Finish. 4. Add the removed static IP address to another interface. a. Select the interface to which you want to move the IP address. b. In the Enabled column, select No to disable the interface. c. Click Configure. d. Set the IP Address to the match the static IP address you removed. e. Click Next, and click Finish. f. In the Enabled column, select Yes to enable the updated interface.
Virtual interface configuration guidelines Virtual interface configuration guidelines apply to failover and aggregate virtual interfaces. There are additional guidelines that apply to either failover or aggregate interfaces but not both. l The virtual-name must be in the form vethx where x is a number. The recommended maximum number is 99 because of name size limitations. l You can create as many virtual interfaces as there are physical interfaces. l Each interface used in a virtual interface must first be disabled. An interface that is part of a virtual interface is seen as disabled for other network configuration options. l After a virtual interface is destroyed, the physical interfaces associated with it remain disabled. You must manually re-enable the physical interfaces. l The number and type of cards installed determines the number of Ethernet ports available. l Each physical interface can belong to one virtual interface. l A system can support multiple mixed failover and aggregation virtual interfaces, subject to the restrictions above. l Virtual interfaces must be created from identical physical interfaces. For example, all copper, all optical, all 1 Gb, or all 10 Gb. However, 1 Gb interfaces support bonding a mix of copper and optical interfaces. This applies to virtual interfaces across different cards with identical physical interfaces, except for Chelsio cards. For Chelsio cards, only failover is supported, and that is only across interfaces on the same card. l Failover and aggregate links improve network performance and resiliency by using two or more network interfaces in parallel, thus increasing the link speed and reliability over that of a single interface.
Network interface management 63 Managing Data Domain Systems
l Remove functionality is available using the Configure button. Click a virtual interface in the list of interfaces on the Interfaces tab and click Configure. From the list of interfaces in the dialog box, clear the checkbox for the interface to remove it from bonding (failover or aggregate), and click Next. l On DD4200, DD4500, and DD7200 systems, the ethMa interface does not support failover or link aggregation. Guidelines for configuring a virtual interface for link aggregation Link aggregation provides improved network performance and resiliency by using one or more network interfaces in parallel, thus increasing the link speed and reliability over that of a single interface. These guidelines are provided to help you optimize your use of link aggregation.
l Changes to disabled Ethernet interfaces flush the routing table. It is recommended that you make interface changes only during scheduled maintenance downtime. Afterwards, reconfigure the routing rules and gateways. l Enable aggregation on an existing virtual interface by specifying the physical interfaces and mode and giving it an IP address. l 10 Gb single-port optical Ethernet cards do not support link aggregation. l 1 GbE and 10 GbE interfaces cannot be aggregated together. l Copper and optical interfaces cannot be aggregated together. l On DD4200, DD4500, and DD7200 systems, the ethMA interface does not support link aggregation. Guidelines for configuring a virtual interface for failover Link failover provides improved network stability and performance by identifying backup interfaces that can support network traffic when the primary interface is not operating. These guidelines are provided to help you optimize your use of link failover.
l A primary interface must be part of the failover. If a primary interface removal is attempted from a failover, an error message appears. l When a primary interface is used in a failover configuration, it must be explicitly specified and must also be a bonded interface to the virtual interface. If the primary interface goes down and multiple interfaces are still available, the next interface is randomly selected. l All interfaces in a virtual interface must be on the same physical network. Network switches used by a virtual interface must be on the same physical network. l The recommended number of physical interfaces for failover is greater than one. You can, however, configure one primary interface and one or more failover interfaces, except with the following:
n 10 Gb CX4 Ethernet card, which are restricted to one primary interface and one failover interface from the same card, and n 10 Gb single-port optical Ethernet cards, which cannot be used. l On DD4200, DD4500, and DD7200 systems, the ethMA interface does not support link failover.
Virtual interface creation Create a virtual interface to support link aggregation or failover. The virtual interface serves as a container for the links to be aggregated or associated for failover.
64 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Creating a virtual interface for link aggregation Create a virtual interface for link aggregation to serve as a container to associate the links that participate in aggregation. A link aggregation interface must specify a link bonding mode and may require a hash selection. For example, you might enable link aggregation on virtual interface veth1 to physical interfaces eth1 and eth2 in mode LACP (Link Aggregation Control Protocol) and hash XOR-L2L3. Procedure 1. Select Hardware > Ethernet > Interfaces. 2. In the Interfaces table, disable the physical interface where the virtual interface is to be added by clicking No in the Enabled column. 3. From the Create menu, select Virtual Interface. 4. In the Create Virtual Interface dialog, specify a virtual interface name in the veth box. Enter a virtual interface name in the form vethx, where x is a unique ID (typically one or two digits). A typical full virtual interface name with VLAN and IP Alias is veth56.3999:199. The maximum length of the full name is 15 characters. Special characters are not allowed. Numbers must be between 0 and 4094, inclusively.
5. In the Bonding Type list, select Aggregate.
Registry settings can be different from the bonding configuration. When interfaces are added to the virtual interface, the information is not sent to the bonding module until the virtual interface is given an IP address and brought up. Until that time the registry and the bonding driver configuration are different.
6. In the Mode list, select a bonding mode. Specify the mode that is compatible with the requirements of the system to which the interfaces are directly attached. l Round-robin Transmit packets in sequential order from the first available link through the last in the aggregated group. l Balanced Data is sent over interfaces as determined by the hash method selected. This requires the associated interfaces on the switch to be grouped into an Ether channel (trunk) and given a hash via the Load Balance parameter. l LACP Link Aggregation Control Protocol is similar to Balanced, except that it uses a control protocol that communicates to the other end and coordinates which links within the bond are available for use. LACP provides a kind of heartbeat failover and must be configured at both ends of the link.
7. If you selected Balanced or LACP mode, specify a bonding hash type in the Hash list. Options are: XOR-L2, XOR-L2L3, or XOR-L3L4. XOR-L2 transmits through a bonded interface with an XOR hash of Layer 2 (inbound and outbound MAC addresses). XOR-L2L3 transmits through a bonded interface with an XOR hash of Layer 2 (inbound and outbound MAC addresses) and Layer 3 (inbound and outbound IP addresses).
Network interface management 65 Managing Data Domain Systems
XOR-L3L4 transmits through a bonded interface with an XOR hash of Layer 3 (inbound and outbound IP addresses) and Layer 4 (inbound and outbound ports).
8. To select an interface to add to the aggregate configuration, select the checkbox that corresponds to the interface, and then click Next. The Create virtual interface veth_name dialog appears.
9. Enter an IP address, or enter 0 to specify no IP address. 10. Enter a netmask address or prefix. 11. Specify Speed/Duplex options. The combination of speed and duplex settings define the rate of data transfer through the interface. Select either:
l Autonegotiate Speed/Duplex Select this option to allow the network interface card to autonegotiate the line speed and duplex setting for an interface. l Manually configure Speed/Duplex Select this option to manually set an interface data transfer rate.
n Duplex options are half-duplex or full-duplex. n Speed options listed are limited to the capabilities of the hardware device. Options are 10 Mb, 100 Mb, 1000 Mb, and 10 Gb. n Half-duplex is only available for 10 Mb and 100 Mb speeds. n 1000 Mb and 10 Gb line speeds require full-duplex. n Optical interfaces require the Autonegotiate option. n The 10 GbE copper NIC default is 10 Gb. If a copper interface is set to 1000 Mb or 10 Gb line speed, duplex must be full-duplex.
12. Specify the MTU setting.
l To select the default value (1500), click Default. l To select a different setting, enter the setting in the MTU box. Ensure that all of your network components support the size set with this option.
13. Optionally, select Dynamic DNS Registration option. Dynamic DNS (DDNS) is a protocol that registers local IP addresses on a Domain Name System (DNS) server. In this release, DD System Manager supports Windows mode DDNS. To use UNIX mode DDNS, use the net ddns CLI command. The DDNS must be registered to enable this option.
14. Click Next. The Configure Interface Settings summary page appears. The values listed reflect the new system and interface state.
15. Click Finish and OK. Creating a virtual interface for link failover Create a virtual interface for link failover to serve as a container to associate the links that will participate in failover. The failover-enabled virtual interface represents a group of secondary interfaces, one of which can be specified as the primary. The system makes the primary interface the active
66 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems interface whenever the primary interface is operational. A configurable Down Delay failover option allows you to configure a failover delay in 900 millisecond intervals. The failover delay guards against multiple failovers when a network is unstable. Procedure
1. Select Hardware > Ethernet > Interfaces. 2. In the interfaces table, disable the physical interface to which the virtual interface is to be added by clicking No in the Enabled column. 3. From the Create menu, select Virtual Interface. 4. In the Create Virtual Interface dialog, specify a virtual interface name in the veth box. Enter a virtual interface name in the form vethx, where x is a unique ID (typically one or two digits). A typical full virtual interface name with VLAN and IP Alias is veth56.3999:199. The maximum length of the full name is 15 characters. Special characters are not allowed. Numbers must be between 0 and 4094, inclusively.
5. In the Bonding Type list, select Failover. 6. Select an interface to add to the failover configuration, and click Next. Virtual aggregate interfaces can be used for failover. The Create virtual interface veth_name dialog appears.
7. Enter an IP address, or enter 0 to specify no IP address. 8. Enter a netmask or prefix. 9. Specify the Speed/Duplex options. The combination of speed and duplex settings defines the rate of data transfer through the interface.
l Select Autonegotiate Speed/Duplex to allow the network interface card to autonegotiate the line speed and duplex setting for an interface. l Select Manually configure Speed/Duplex to manually set an interface data- transfer rate.
n Duplex options are either half duplex or full duplex. n Speed options listed are limited to the capabilities of the hardware device. Options are 10 Mb, 100 Mb, 1000 Mb, and 10 Gb. n Half-duplex is available for 10 Mb and 100 Mb speeds only. n 1000 Mb and 10 Gb line speeds require full-duplex. n Optical interfaces require the Autonegotiate option. n The copper interface default is 10 Gb. If a copper interface is set to 1000 Gb or 10 Gb line speed, the duplex must be full-duplex.
10. Specify MTU setting.
l To select the default value (1500), click Default. l To select a different setting, enter the setting in the MTU box. Ensure that all of your network path components support the size set with this option.
11. Optionally, select Dynamic DNS Registration option. Dynamic DNS (DDNS) is a protocol that registers local IP addresses on a Domain Name System (DNS) server. In this release, DD System Manager supports Windows mode DDNS. To use UNIX mode DDNS, use the net ddns CLI command.
Network interface management 67 Managing Data Domain Systems
The DDNS must be registered to enable this option.
This option disables DHCP for this interface.
12. Click Next. The Configure Interface Settings summary page appears. The values listed reflect the new system and interface state.
13. Complete the Interface, click Finish and OK.
Modifying a virtual interface After you create a virtual interface, you can update the settings to respond to network changes or resolve issues. Procedure 1. Select Hardware > Ethernet > Interfaces. 2. In the Interfaces column, select the interface and disable the virtual interface by clicking No in the Enabled column. Click OK in the warning dialog. 3. In the Interfaces column, select the interface and click Configure. 4. In the Configure Virtual Interface dialog, change the settings. 5. Click Next and Finish.
Configuring a VLAN Create a new VLAN interface from either a physical interface or a virtual interface. The recommended total VLAN count is 80. You can create up to 100 interfaces (minus the number of aliases, physical and virtual interfaces) before the system prevents you from creating any more. Procedure 1. Select Hardware > Ethernet > Interfaces. 2. In the interfaces table, select the interface to which you want to add the VLAN. The interface you select must be configured with an IP address before you can add a VLAN. 3. Click Create and selectVLAN. 4. In the Create VLAN dialog box, specify a VLAN ID by entering a number in the VLAN Id box. The range of a VLAN ID is between 1 and 4094 inclusive.
5. Enter an IP address, or enter 0 to specify no IP address. The Internet Protocol (IP) address is the numerical label assigned to the interface. For example, 192.168.10.23.
6. Enter a netmask or prefix. 7. Specify the MTU setting. The VLAN MTU must be less than or equal to the MTU defined for the physical or virtual interface to which it is assigned. If the MTU defined for the supporting physical or virtual interface is reduced below the configured VLAN value, the VLAN value is
68 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
automatically reduced to match the supporting interface. If the MTU value for the supporting interface is increased above the configured VLAN value, the VLAN value is unchanged.
l To select the default value (1500), click Default. l To select a different setting, enter the setting in the MTU box. DD System Manager does not accept an MTU size that is larger than that defined for the physical or virtual interface to which the VLAN is assigned.
8. Specify Dynamic DNS Registration option. Dynamic DNS (DDNS) is a protocol that registers local IP addresses on a Domain Name System (DNS) server. In this release, DD System Manager supports Windows mode DDNS. To use UNIX mode DDNS, use the net ddns CLI command. The DDNS must be registered to enable this option.
9. Click Next. The Create VLAN summary page appears.
10. Review the configuration settings, click Finish, and click OK.
Modifying a VLAN interface After you create a VLAN interface, you can update the settings to respond to network changes or resolve issues. Procedure
1. Select Hardware > Ethernet > Interfaces. 2. In the Interfaces column, select the checkbox of the interface and disable the VLAN interface by clicking No in the Enabled column. Click OK in the warning dialog box. 3. In the Interfaces column, select the checkbox of the interface and click Configure. 4. In the Configure VLAN Interface dialog, change the settings. 5. Click Next and Finish.
Configuring an IP alias An IP alias assigns an additional IP address to a physical interface, a virtual interface, or a VLAN. The recommended total number of IP aliases, VLAN, physical, and virtual interfaces that can exist on the system is 80. Although up to 100 interfaces are supported, as the maximum number is approached, you might notice slowness in the display.
When using a Data Domain HA system, if a user is created and logins to the standby node without logging into the active node first, the user will not have a default alias to use. Therefore, in order to use aliases on the standby node, the user should login to the active node first.
1. Select Hardware > Ethernet > Interfaces. 2. Click Create, and select IP Alias. The Create IP Alias dialog appears.
Network interface management 69 Managing Data Domain Systems
3. Specify an IP alias ID by entering a number in the IP ALIAS Id box. The range is 1 to 4094 inclusive.
4. Enter an IPv4 or IPv6 address. 5. If you entered an IPv4 address, enter a netmask address. 6. Specify Dynamic DNS Registration option. Dynamic DNS (DDNS) is a protocol that registers local IP addresses on a Domain Name System (DNS) server. In this release, DD System Manager supports Windows mode DDNS. To use UNIX mode DDNS, use the net ddns CLI command. The DDNS must be registered to enable this option.
7. Click Next. The Create IP Alias summary page appears.
8. Review the configuration settings, click Finish, and OK.
Modifying an IP alias interface After you create an IP alias, you can update the settings to respond to network changes or resolve issues. Procedure
1. Select Hardware > Ethernet > Interfaces. 2. In the Interfaces column, select the checkbox of the interface and disable the IP alias interface by clicking No in the Enabled column. Click OK in the warning dialog box. 3. In the Interfaces column, select the checkbox of the interface and click Configure. 4. In the Configure IP Alias dialog box, change the settings as described in the procedure for creating an IP Alias. 5. Click Next and Finish.
Registering interfaces with DDNS Dynamic DNS (DDNS) is a protocol that registers local IP addresses on a Domain Name System (DNS) server. In this release, DD System Manager supports Windows mode DDNS. To use UNIX mode DDNS, use the net ddns CLI command. You can do the following.
l Manually register (add) configured interfaces to the DDNS registration list. l Remove interfaces from the DDNS registration list. l Enable or disable DNS updates. l Display whether DDNS registration is enabled or not. l Display interfaces in the DDNS registration list. Procedure
1. Select Hardware > Ethernet > Interfaces > DDNS Registration. 2. In the DDNS Windows Mode Registration dialog, click Add to add an interface to the DDNS. The Add Interface dialog box appears.
70 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
a. Enter a name in the Interface field. b. Click OK. 3. Optionally, to remove an interface from the DDNS: a. Select the interface to remove, and click Remove. b. In the Confirm Remove dialog box, click OK. 4. Specify the DDNS Status.
l Select Enable to enable updates for all interfaces already registered. l Click Default to select the default settings for DDNS updates. l Clear Enable to disable DDNS updates for the registered interfaces.
5. To complete the DDNS registration, click OK.
Destroying an interface You can use DD System Manager to destroy or delete virtual, VLAN, and IP alias interfaces. When a virtual interface is destroyed, the system deletes the virtual interface, releases its bonded physical interface, and deletes any VLANs or aliases attached to the virtual interface. When you delete a VLAN interface, the OS deletes the VLAN and any IP alias interfaces that are created under it. When you destroy an IP alias, the OS deletes only that alias interface. Procedure
1. Select Hardware > Ethernet > Interfaces. 2. Click the box next to each interface you want to destroy (Virtual or VLAN or IP Alias). 3. Click Destroy. 4. Click OK to confirm.
Viewing an interface hierarchy in the tree view The Tree View dialog displays the association between physical and virtual interfaces. Procedure
1. Select Hardware > Ethernet > Interfaces > Tree View. 2. In the Tree View dialog box, click the plus or minus boxes to expand or contract the tree view that shows the hierarchy. 3. Click Close to exit this view.
General network settings management The configuration settings for hostname, domain name, search domains, host mapping, and DNS list are managed together on the Settings tab.
Viewing network settings information The Settings tab displays the current configuration for the hostname, domain name, search domains, host mapping, and DNS. Procedure
1. Select Hardware > Ethernet > Settings.
General network settings management 71 Managing Data Domain Systems
Results The Settings tab displays the following information. Host Settings Host Name The hostname of the selected system. Domain Name The fully qualified domain name associated with the selected system.
Search Domain List Search Domain A list of search domains that the selected system uses. The system applies the search domain as a suffix to the hostname.
Hosts Mapping IP Address IP address of the host to resolve. Host Name Hostnames associated with the IP address.
DNS List DNS IP Address Current DNS IP addresses associated with the selected system. An asterisk (*) indicates that the IP addresses were assigned through DHCP.
Setting the DD System Manager hostname You can configure the DD System Manager hostname and domain name manually, or you can configure DD OS to automatically receive the host and domain names from a Dynamic Host Configuration Protocol (DHCP) server. One advantage to manually configuring the host and domain names is that you remove the dependency on the DHCP server and the interface leading to the DHCP server. To minimize the risk of service interruption, EMC recommends that you manually configure the host and domain names. When configuring the hostname and domain name, consider the following guidelines.
l Do not include an underscore in the hostname; it is incompatible with some browsers. l Replication and CIFS authentication must be reconfigured after you change the names. l If a system was previously added without a fully qualified name (no domain name), a domain name change requires that you remove and add the affected system or update the Search Domain List to include the new domain name. Procedure
1. Select Hardware > Ethernet > Settings. 2. Click Edit in the Host Settings area. The Configure Host dialog appears.
72 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
3. To manually configure the host and domain names: a. Select Manually configure host. b. Enter a hostname in the Host Name box.
For example, id##.yourcompany.com c. Enter a domain name in the Domain Name box. This is the domain name associated with your Data Domain system and, usually, your company’s domain name. For example, yourcompany.com
d. Click OK. The system displays progress messages as the changes are applied. 4. To obtain the host and domain names from a DHCP server, select Obtain Settings using DHCP and click OK. At least one interface must be configured to use DHCP.
Managing the domain search list Use the domain search list to define which domains the system can search. Procedure
1. Select Hardware > Ethernet > Settings. 2. Click Edit in the Search Domain List area. 3. To add a search domain using the Configure Search Domains dialog: a. Click Add (+). b. In the Add Search Domain dialog, enter a name in the Search Domain box.
For example, id##.yourcompany.com c. Click OK. The system adds the new domain to the list of searchable domains. d. Click OK to apply changes and return to the Settings view. 4. To remove a search domain using the Configure Search Domains dialog: a. Select the search domain to remove. b. Click Delete (X). The system removes the selected domain from the list of searchable domains. c. Click OK to apply changes and return to the Settings view.
Adding and deleting host maps A host map links an IP address to a hostname, so that either the IP address or the hostname can be used to specify the host. Procedure
1. Select Hardware > Ethernet > Settings. 2. To add a host map, do the following. a. In the Hosts Mapping area, click Add. b. In the Add Hosts dialog, enter the IP address of the host in the IP Address box.
General network settings management 73 Managing Data Domain Systems
d. In the Add Host dialog, enter a hostname, such as id##.yourcompany.com, in the Host Name box. e. Click OK to add the new hostname to the Host Name list. f. Click OK to return to the Settings tab. 3. To delete a host map, do the following. a. In the Hosts Mapping area, select the host mapping to delete. b. Click Delete (X).
Configuring DNS IP addresses DNS IP addresses specify the DNS servers the system can use to get IP addresses for host names that are not in the host mapping table. You can configure the DNS IP addresses manually, or you can configure DD OS to automatically receive IP addresses from a DHCP server. One advantage to manually configuring DNS IP addresses is that you remove the dependency on the DHCP server and the interface leading to the DHCP server. To minimize the risk of service interruption, EMC recommends that you manually configure the DNS IP addresses. Procedure
1. Select Hardware > Ethernet > Settings. 2. Click Edit in the DNS List area. 3. To manually add a DNS IP address: a. Select Manually configure DNS list. The DNS IP address checkboxes become active. b. Click Add (+). c. In the Add DNS dialog box, enter the DNS IP address to add. d. Click OK. The system adds the new IP address to the list of DNS IP addresses. e. Click OK to apply the changes. 4. To delete a DNS IP address from the list: a. Select Manually configure DNS list. The DNS IP address checkboxes become active. b. Select the DNS IP address to delete and click Delete (X). The system removes the IP address from the list of DNS IP addresses. c. Click OK to apply the changes. 5. To obtain DNS addresses from a DHCP server, select Obtain DNS using DHCP and click OK. At least one interface must be configured to use DHCP.
74 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Network route management Routes determine the path taken to transfer data to and from the localhost (the Data Domain system) to another network or host. Data Domain systems do not generate or respond to any of the network routing management protocols (RIP, EGRP/EIGRP, and BGP). The only routing implemented on a Data Domain system is based upon the internal route table, in which the administrator may define a specific network or subnet that a physical interface (or interface group) uses. Data Domain systems use source-based routing, which means that outbound network packets that match the subnet of multiple interfaces are routed only over the physical interface from which they originated. Set static routes multiple interfaces contain the same IPv6 subnets, and the connections are being made to IPv6 addresses with this subnet. Normally, static routes are not needed with IPv4 addresses with the same subnet, such as for backups. There are cases in which static addresses may be required to allow connections to work, such as connections from the Data Domain system to remote systems.
Routing for connections initiated from the Data Domain system, such as for replication, depends on the source address used for interfaces on the same subnet. To force traffic for a specific interface to a specific destination (even if that interface is on the same subnet as other interfaces), configure a static routing entry between the two systems: this static routing overrides source routing.
Viewing route information The Routes tab displays the default gateways, static routes, and dynamic routes. Procedure
1. Select Hardware > Ethernet > Routes. Results The Static Routes area lists the route specification used to configure each static route. The Dynamic Routes table lists information for each of the dynamically assigned routes.
Table 24 Dynamic Routes column label descriptions
Destination The destination host/network where the network traffic (data) is sent. Gateway The address of the router in the DD network, or 0.0.0.0 if no gateway is set. Genmask The netmask for the destination net. Set to 255.255.255.255 for a host destination and 0.0.0.0 for the default route. Flags Possible flags include: U—Route is up, H—Target is a host, G —Use gateway, R —Reinstate route for dynamic routing, D—Dynamically installed by daemon or redirect, M —Modified from routing daemon or redirect, A — Installed by addrconf, C —Cache entry, and ! —Reject route. Metric The distance to the target (usually counted in hops). Not used by the DD OS, but might be needed by routing daemons.
Network route management 75 Managing Data Domain Systems
Table 24 Dynamic Routes column label descriptions (continued)
MTU Maximum Transfer Unit (MTU) size for the physical (Ethernet) interface. Window Default window size for TCP connections over this route. IRTT Initial RTT (Round Trip Time) used by the kernel to estimate the best TCP protocol parameters without waiting on possibly slow answers. Interface Interface name associated with the routing interface.
Setting the default gateway You can configure the default gateway manually, or you can configure DD OS to automatically receive the default gateway IP addresses from a DHCP server. One advantage to manually configuring the default gateway is that you remove the dependency on the DHCP server and the interface leading to the DHCP server. To minimize the risk of service interruption, EMC recommends that you manually configure the default gateway IP address. Procedure
1. Select Hardware > Ethernet > Routes. 2. Click Edit next to the default gateway type (IPv4 or IPv6) you want to configure. 3. To manually configure the default gateway address: a. Select Manually Configure. b. Enter the gateway address in the Gateway box. c. Click OK. 4. To obtain the default gateway address from a DHCP server, select Use DHCP value and click OK. At least one interface must be configured to use DHCP.
Creating static routes Static routes define destination hosts or networks that they system can communicate with. Procedure
1. Select Hardware > Ethernet > Routes. 2. Click Create in the Static Routes area. 3. In the Create Routes dialog, select the interface you want to host the static route, and click Next. 4. Specify the Destination.
l To specify a destination network, select Network and enter the network address and netmask for the destination network. l To specify a destination host, select Host and enter the hostname or IP address of the destination host.
5. Optionally, specify the gateway to use to connect to the destination network or host.
76 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
a. Select Specify a gateway for this route. b. Enter the gateway address in the Gateway box. 6. Review the configuration and click Next. The create routes Summary page appears.
7. Click Finish. 8. After the process is completed, click OK. The new route specification is listed in the Route Spec list.
Deleting static routes Delete a static route when you no longer want the system to communicate with a destination host or network. Procedure 1. Select Hardware > Ethernet > Routes. 2. Select the Route Spec of the route specification to delete. 3. Click Delete. 4. Click Delete to confirm and then click Close. The selected route specification is removed from the Route Spec list.
System passphrase management The system passphrase is a key that allows a Data Domain system to be transported with encryption keys on the system. The encryption keys protect the data and the system passphrase protects the encryption keys. The system passphrase is a human-readable (understandable) key (like a smart card) which is used to generate a machine usable AES 256 encryption key. If the system is stolen in transit, an attacker cannot easily recover the data; at most, they can recover the encrypted user data and the encrypted keys. The passphrase is stored internally on a hidden part the Data Domain storage subsystem. This allows the Data Domain system to boot and continue servicing data access without any administrator intervention.
Setting the system passphrase The system passphrase must be set before the system can support data encryption or request digital certificates. Before you begin No minimum system passphrase length is configured when DD OS is installed, but the CLI provides a command to set a minimum length. To determine if a minimum length is configured for the passphrase, enter the system passphrase option show CLI command. Procedure 1. Select Administration > Access > Administrator Access. If the system passphrase is not set, the Set Passphrase button appears in the Passphrase area. If a system passphrase is configured, the Change Passphrase button appears, and your only option is to change the passphrase.
System passphrase management 77 Managing Data Domain Systems
2. Click the Set Passphrase button. The Set Passphrase dialog appears.
3. Enter the system passphrase in the boxes and click Next. If a minimum length is configured for the system passphrase, the passphrase you enter must contain the minimum number of characters.
Results The system passphrase is set and the Change Passphrase button replaces the Set Passphrase button.
Changing the system passphrase The administrator can change the passphrase without having to manipulate the actual encryption keys. Changing the passphrase indirectly changes the encryption of the keys, but does not affect user data or the underlying encryption key. Changing the passphrase requires two-user authentication to protect against data shredding. Procedure
1. Select Administration > Access > Administrator Access. 2. To change the system passphrase, click Change Passphrase. The Change Passphrase dialog appears.
The file system must be disabled to change the passphrase. If the file system is running, you are prompted to disable it.
3. In the text fields, provide:
l The user name and password of a Security Officer account (an authorized user in the Security User group on that Data Domain system). l The current passphrase when changing the passphrase. l The new passphrase, which must contain the minimum number of characters configured with the system passphrase option set min-length command.
4. Click the checkbox for Enable file system now. 5. Click OK.
Be sure to take care of the passphrase. If the passphrase is lost, you can never unlock the file system and access the data; the data is irrevocably lost.
78 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
System access management System access management features allow you to control system access to users in a local database or in a network directory. Additional controls define different access levels and control which protocols can access the system.
Role-based access control Role-based access control (RBAC) is an authentication policy that controls which DD System Manager controls and CLI commands a user can access on a system. For example, users who are assigned the admin role can configure and monitor an entire system, while users who are assigned the user role are limited to monitoring a system. When logged into DD System Manager, users see only the program features that they are permitted to use based on the role assigned to the user. The following roles are available for administering and managing the DD OS. admin An admin role user can configure and monitor the entire Data Domain system. Most configuration features and commands are available only to admin role users. However, some features and commands require the approval of a security role user before a task is completed. user The user role enables users to monitor systems and change their own password. Users who are assigned the user management role can view system status, but they cannot change the system configuration. security (security officer) A security role user, who may be referred to as a security officer, can manage other security officers, authorize procedures that require security officer approval, and perform all tasks supported for user-role users. The security role is provided to comply with the Write Once Read-Many (WORM) regulation. This regulation requires electronically stored corporate data be kept in an unaltered, original state for purposes such as eDiscovery. EMC Data Domain added auditing and logging capabilities to enhance this feature. As a result of compliance regulations, most command options for administering sensitive operations, such as encryption, Retention Lock Compliance, and archiving now require security officer approval. In a typical scenario, an admin role user issues a command and, if security officer approval is required, the system displays a prompt for approval. To proceed with the original task, the security officer must enter his or her username and password on the same console at which the command was run. If the system recognizes the security officer credentials, the procedure is authorized. If not, a security alert is generated. The following are some guidelines that apply to security-role users:
l Only the sysadmin user (the default user created during the DD OS installation) can create the first security officer, after which the privilege to create security officers is removed from the sysadmin user.
l After the first security officer is created, only security officers can create other security officers.
System access management 79 Managing Data Domain Systems
l Creating a security officer does not enable the authorization policy. To enable the authorization policy, a security officer must log in and enable the authorization policy.
l Separation of privilege and duty apply. admin role users cannot perform security officer tasks, and security officers cannot perform system configuration tasks.
l During an upgrade, if the system configuration contains security officers, a sec- off-defaults permission is created that includes a list of all current security officers.
backup-operator A backup-operator role user can perform all tasks permitted for user role users, create snapshots for MTrees, import, export, and move tapes between elements in a virtual tape library, and copy tapes across pools. A backup-operator role user can also add and delete SSH public keys for non- password-required log ins. (This function is used mostly for automated scripting.) He or she can add, delete, reset and view CLI command aliases, synchronize modified files, and wait for replication to complete on the destination system. none The none role is for DD Boost authentication and tenant-unit users only. A none role user can log in to a Data Domain system and can change his or her password, but cannot monitor, manage, or configure the primary system. When the primary system is partitioned into tenant units, either the tenant-admin or the tenant-user role is used to define a user's role with respect to a specific tenant unit. The tenant user is first assigned the none role to minimize access to the primary system, and then either the tenant-admin or the tenant-user role is appended to that user. tenant-admin A tenant-admin role can be appended to the other (non-tenant) roles when the Secure Multi-Tenancy (SMT) feature is enabled. A tenant-admin user can configure and monitor a specific tenant unit. tenant-user A tenant-user role can be appended to the other (non-tenant) roles when the SMT feature is enabled. The tenant-user role enables a user to monitor a specific tenant unit and change the user password. Users who are assigned the tenant-user management role can view tenant unit status, but they cannot change the tenant unit configuration.
Access management for IP protocols This feature manages system access for the FTP, FTPS, HTTP, HTTPS, SSH, SCP, and Telnet protocols.
Viewing the IP services configuration The Administrator Access tab displays the configuration status for the IP protocols that can be used to access the system. FTP and FTPS are the only protocols that are restricted to administrators. Procedure
1. Select Administration > Access > Administrator Access.
80 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Results The Access Management page displays the Administrator Access, Local Users, Authentication, and Active Users tabs.
Table 25 Administrator Access tab information
Passphrase If no passphrase is set, the Set Passphrase button appears. If a passphrase is set, the Change Passphrase button appears. Services The name of a service/protocol that can access the system. Enabled (Yes/No) The status of the service. If the service is disabled, enable it by selecting it in the list and clicking Configure. Fill out the General tab of the dialog box. If the service is enabled, modify its settings by selecting it in the list and clicking Configure. Edit the settings in the General tab of the dialog box.
Allowed Hosts The host or hosts that can access the service. Service Options The port or session timeout value for the service selected in the list. FTP/FTPS Only the session timeout can be set. HTTP port The port number opened for the HTTP protocol (port 80, by default). HTTPS port The port number opened for the HTTPS protocol (port 443, by default). SSH/SCP port The port number opened for the SSH/SCP protocol (port 22, by default). Telnet No port number can be set. Session Timeout The amount of inactive time allowed before a connection closes. The default is Infinite, that is, the connection does not close. EMC recommends a session timeout maximum of five minutes. Use the Advanced tab of the dialog box to set a timeout in seconds.
Managing FTP access The File Transfer Protocol (FTP) allows administrators to access files on the Data Domain system. You can enable either FTP or FTPS access to users who are assigned the admin management role. FTP access allows admin user names and passwords to cross the network in clear text, making FTP an insecure access method. FTPS is recommended as a secure access method. When you enable either FTP or FTPS access, the other access method is disabled.
Only users who are assigned the admin management role are permitted to access the system using FTP
LFTP clients that connect to a Data Domain system via FTPS or FTP are disconnected after reaching a set timeout limit. However the LFTP client uses its cached username and password to reconnect after the timeout while you are running any command.
Access management for IP protocols 81 Managing Data Domain Systems
1. Select Administration > Access > Administrator Access. 2. Select FTP and click Configure. 3. To manage FTP access and which hosts can connect, select the General tab and do the following: a. To enable FTP access, select Allow FTP Access. b. To enable all hosts to connect, select Allow all hosts to connect. c. To restrict access to select hosts, select Limit Access to the following systems, and modify the Allowed Hosts list.
You can identify a host using a fully qualified hostname, an IPv4 address, or an IPv6 address.
l To add a host, click Add (+). Enter the host identification and click OK. l To modify a host ID, select the host in the Hosts list and click Edit (pencil). Change the host ID and click OK. l To remove a host ID, select the host in the Hosts list and click Delete (X).
4. To set a session timeout, select the Advanced tab, and enter the timeout value in seconds.
The session timeout default is Infinite, that is, the connection does not close.
5. Click OK. If FTPS is enabled, a warning message appears with a prompt to click OK to proceed.
Managing FTPS access The FTP Secure (FTPS) protocol allows administrators to access files on the Data Domain system. FTPS provides additional security over using FTP, such as support for the Transport Layer Security (TLS) and for the Secure Sockets Layer (SSL) cryptographic protocols. Consider the following guidelines when using FTPS.
l Only users who are assigned the admin management role are permitted to access the system using FTPS. l When you enable FTPS access, FTP access is disabled. l FTPS does not show up as a service for DD systems that run DD OS 5.2, managed from a DD system running DD OS 5.3 or later. l When you issue the get command, the fatal error message SSL_read: wrong version number lftp appears if matching versions of SSL are not installed on the Data Domain system and compiled on the LFTP client . As a workaround, attempt to re-issue the get command on the same file. Procedure
1. Select Administration > Access > Administrator Access.
82 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
2. Select FTPS and click Configure. 3. To manage FTPS access and which hosts can connect, select the General tab and do the following: a. To enable FTPS access, select Allow FTPS Access. b. To enable all hosts to connect, select Allow all hosts to connect. c. To restrict access to select hosts, select Limit Access to the following systems, and modify the hosts list.
You can identify a host using a fully qualified hostname, an IPv4 address, or an IPv6 address.
l To add a host, click Add (+). Enter the host identification and click OK. l To modify a host ID, select the host in the Hosts list and click Edit (pencil). Change the host ID and click OK. l To remove a host ID, select the host in the Hosts list and click Delete (X).
4. To set a session timeout, select the Advanced tab and enter the timeout value in seconds.
The session timeout default is Infinite, that is, the connection does not close.
5. Click OK. If FTP is enabled, a warning message appears and prompts you to click OK to proceed.
Managing HTTP and HTTPS access HTTP or HTTPS access is required to support browser access to DD System Manager. Procedure 1. Select Administration > Access > Administrator Access. 2. Select HTTP or HTTPS and click Configure. The Configure HTTP/HTTPS Access dialog appears and displays tabs for general configuration, advanced configuration, and certificate management. 3. To manage the access method and which hosts can connect, select the General tab and do the following: a. Select the checkboxes for the access methods you want to allow. b. To enable all hosts to connect, select Allow all hosts to connect. c. To restrict access to select hosts, select Limit Access to the following systems, and modify the host list.
You can identify a host using a fully qualified hostname, an IPv4 address, or an IPv6 address.
l To add a host, click Add (+). Enter the host identification and click OK. l To modify a host ID, select the host in the Hosts list and click Edit (pencil). Change the host ID and click OK.
Access management for IP protocols 83 Managing Data Domain Systems
l To remove a host ID, select the host in the Hosts list and click Delete (X).
4. To configure system ports and session timeout values, select the Advanced tab, and complete the form.
l In the HTTP Port box, enter the port number. Port 80 is assigned by default. l In the HTTPS Port box, enter the number. Port 443 is assigned by default. l In the Session Timeout box, enter the interval in seconds that must elapse before a connection closes. The minimum is 60 seconds and the maximum is 31536000 seconds (one year).
The session timeout default is Infinite, that is, the connection does not close.
Managing host certificates for HTTP and HTTPS A host certificate allows browsers to verify the identity of the system when establishing management sessions.
Requesting a host certificate for HTTP and HTTPS You can use DD System Manager to generate a host certificate request, which you can then forward to a Certificate Authority (CA).
You must configure a system passphrase (system passphrase set) before you can generate a CSR.
1. Select Administration > Access > Administrator Access. 2. In the Services area, select HTTP or HTTPSand click Configure. 3. Select the Certificate tab. 4. Click Add. A dialog appears for the protocol you selected earlier in this procedure.
5. Click Generate the CSR for this Data Domain system. The dialog expands to display a CSR form.
DD OS supports one active CSR at a time. After a CSR is generated, the Generate the CSR for this Data Domain system link is replaced with the Download the CSR for this Data Domain system link. To delete a CSR, use the adminaccess certificate cert-signing-request delete CLI command.
6. Complete the CSR form and click Generate and download a CSR. The CSR file is saved at the following path: /ddvar/certificates/ CertificateSigningRequest.csr. Use SCP, FTP or FTPS to transfer the CSR file from the system to a computer from which you can send the CSR to a CA.
84 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Adding a host certificate for HTTP and HTTPS You can use DD System Manager to add a host certificate to the system. Procedure 1. If you did not requested a host certificate, request a host certificate from a certificate authority. 2. When you receive a host certificate, copy or move it to the computer from which you run DD Service Manager. 3. Select Administration > Access > Administrator Access. 4. In the Services area, select HTTP or HTTPS and click Configure. 5. Select the Certificate tab. 6. Click Add. A dialog appears for the protocol you selected earlier in this procedure.
7. To add a host certificate enclosed in a .p12 file, do the following: a. Select I want to upload the certificate as a .p12 file. b. Type the password in the Password box. c. Click Browse and select the host certificate file to upload to the system. d. Click Add. 8. To add a host certificate enclosed in a .pem file, do the following: a. Select I want to upload the public key as a .pem file and use a generated private key. b. Click Browse and select the host certificate file to upload to the system. c. Click Add. Deleting a host certificate for HTTP and HTTPS DD OS supports one host certificate for HTTP and HTTPS. If the system is currently using a host certificate and you want to use a different host certificate, you must delete the current certificate before adding the new certificate. Procedure
1. Select Administration > Access > Administrator Access. 2. In the Services area, select HTTP or HTTPS and click Configure. 3. Select the Certificate tab. 4. Select the certificate you want to delete. 5. Click Delete, and click OK.
Managing SSH and SCP access SSH is a secure protocol that enables network access to the system CLI, with or without SCP (secure copy). You can use DD System Manager to enable system access using the SSH protocol. SCP requires SSH, so when SSH is disabled, SCP is automatically disabled. Procedure
1. Select Administration > Access > Administrator Access. 2. Select SSH or SCP and click Configure.
Access management for IP protocols 85 Managing Data Domain Systems
3. To manage the access method and which hosts can connect, select the General tab. a. Select the checkboxes for the access methods you want to allow. b. To enable all hosts to connect, select Allow all hosts to connect. c. To restrict access to select hosts, select Limit Access to the following systems, and modify the host list.
You can identify a host using a fully qualified hostname, an IPv4 address, or an IPv6 address.
l To add a host, click Add (+). Enter the host identification and click OK. l To modify a host ID, select the host in the Hosts list and click Edit (pencil). Change the host ID and click OK. l To remove a host ID, select the host in the Hosts list and click Delete (X).
4. To configure system ports and session timeout values, click the Advanced tab.
l In the SSH/SCP Port text entry box, enter the port number. Port 22 is assigned by default. l In the Session Timeout box, enter the interval in seconds that must elapse before connection closes.
The session timeout default is Infinite, that is, the connection does not close.
Click Default to revert to the default value.
Managing Telnet access Telnet is an insecure protocol that enables network access to the system CLI.
Telnet access allows user names and passwords to cross the network in clear text, making Telnet an insecure access method.
1. Select Administration > Access > Administrator Access. 2. Select Telnet and click Configure. 3. To manage Telnet access and which hosts can connect, select the General tab. a. To enable Telnet access, select Allow Telnet Access. b. To enable all hosts to connect, select Allow all hosts to connect. c. To restrict access to select hosts, select Limit Access to the following systems, and modify the host list.
86 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
You can identify a host using a fully qualified hostname, an IPv4 address, or an IPv6 address.
l To add a host, click Add (+). Enter the host identification and click OK. l To modify a host ID, select the host in the Hosts list and click Edit (pencil). Change the host ID and click OK. l To remove a host ID, select the host in the Hosts list and click Delete (X).
4. To set a session timeout, select the Advanced tab and enter the timeout value in seconds.
The session timeout default is Infinite, that is, the connection does not close.
Local user account management A local user is a user account (user name and password) that is configured on the Data Domain system instead of being defined in a Windows Active Directory, Windows Workgroup, or NIS directory.
UID conflicts: local user and NIS user accounts When you set up a Data Domain system in an NIS environment, be aware of potential UID conflicts between local and NIS user accounts. Local user accounts on a Data Domain system start with a UID of 500. To avoid conflicts, consider the size of potential local accounts when you define allowable UID ranges for NIS users.
Viewing local user information Local users are user accounts that are defined on the system, rather than in Active Directory, a Workgroup, or UNIX. You can display the local user's username, management role, login status, and target disable date. You can also display the user's password controls and the tenant units the user can access.
The user-authentication module uses Greenwich Mean Time (GMT). To ensure that user accounts and passwords expire correctly, configure settings to use the GMT that corresponds to the target local time.
1. Select Administration > Access > Local Users . The Local Users view appears and shows the Local Users table and the Detailed Information area.
Local user account management 87 Managing Data Domain Systems
Table 26 Local user list column label descriptions
Name The user ID, as added to the system. Management Role The role displayed is admin, user, security, backup-operator, or none. In this table, Tenant user roles are displayed as none. To see an assigned tenant role, select the user and view the role in the Detailed Information area.
Status l Active—User access to the account is permitted.
l Disabled—User access to the account is denied because the account is administratively disabled, the current date is beyond the account expiration date, or a locked account’s password requires renewal.
l Locked—User access is denied because the password expired.
Disable Date The date the account is set to be disabled. Last Login From The location where the user last logged in. Last Login Time The time the user last logged in.
User accounts configured with the admin or security officer roles can view all users. Users with other roles can view only their own user accounts.
2. Select the user you want to view from the list of users. Information about the selected user displays in the Detailed Information area.
Table 27 Detailed User Information, Row Label Descriptions
Tenant-User The list of tenant units the user can access as a tenant-user role user. Tenant-Admin The list of tenant units the user can access as a tenant-admin role user. Password Last Changed The date the password was last changed. Minimum Days Between The minimum number of days between password changes that you Change allow a user. Default is 0. Maximum Days Between The maximum number of days between password changes that you Change allow a user. Default is 90. Warn Days Before Expire The number of days to warn the users before their password expires. Default is 7. Disable Days After Expire The number of days after a password expires to disable the user account. Default is Never.
88 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
The default values are the initial default password policy values. A system administrator (admin role) can change them by selecting More Tasks > Change Login Options.
Creating local users Create local users when you want to manage access on the local system instead of through an external directory. Data Domain systems support a maximum of 500 local user accounts. Procedure
1. Select Administration > Access > Local Users. The Local Users view appears.
2. Click Create to create a new user. The Create User dialog appears.
3. Enter user information in the General Tab.
Table 28 Create User dialog, general controls
User The user ID or name. Password The user password. Set a default password, and the user can change it later. Verify Password The user password, again. Management Role The role assigned to the user, which can be admin, user, security, backup-operator, or none. .
Only the sysadmin user (the default user created during the DD OS installation) can create the first security-role user. After the first security-role user is created, only security-role users can create other security-role users.
Force Password Change Select this checkbox to require that the user change the password during the first login when logging in to DD System Manager or to the CLI with SSH or Telnet.
The default value for the minimum length of a password is 6 characters. The default value for the minimum number of character classes required for a user password is 1. Allowable character classes include:
l Lowercase letters (a-z) l Uppercase letters (A-Z) l Numbers (0-9) l Special Characters ($, %, #, +, and so on)
Local user account management 89 Managing Data Domain Systems
Sysadmin is the default admin-role user and cannot be deleted or modified.
4. To manage password and account expiration, select the Advanced tab and use the controls described in the following table.
Table 29 Create User dialog, advanced controls
Minimum Days Between The minimum number of days between password changes that you Change allow a user. Default is 0. Maximum Days Between The maximum number of days between password changes that you Change allow a user. Default is 90. Warn Days Before Expire The number of days to warn the users before their password expires. Default is 7.
Disable Days After Expire The number of days after a password expires to disable the user account. Default is Never. Disable account on the Check this box and enter a date (mm/dd/yyyy) when you want to following date disable this account. Also, you can click the calendar to select a date.
Note: The default password policy can change if an admin-role user changes them (More Tasks > Change Login Options). The default values are the initial default password policy values.
Modifying a local user profile After you create a user, you can use DD System Manager to modify the user configuration. Procedure
1. Select Administration > Access > Local Users. The Local Users view appears.
2. Click a user name from the list. 3. Click Modify to make changes to a user account. The Modify User dialog box appears. 4. Update the information on the General tab.
If SMT is enabled and a role change is requested from none to any other role, the change is accepted only if the user is not assigned to a tenant-unit as a management- user, is not a DD Boost user with its default-tenant-unit set, and is not the owner of a storage-unit that is assigned to a tenant-unit.
90 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 30 Modify User dialog, general controls
User The user ID or name. Role Select the role from the list.
5. Update the information on the Advanced tab.
Table 31 Modify User dialog, advanced controls
Minimum Days Between The minimum number of days between password changes that you Change allow a user. Default is 0.
Maximum Days Between The maximum number of days between password changes that you Change allow a user. Default is 90. Warn Days Before Expire The number of days to warn the users before their password expires. Default is 7. Disable Days After Expire The number of days after a password expires to disable the user account. Default is Never.
Deleting a local user You can delete certain users based on your user role. If one of the selected users cannot be deleted, the Delete button is disabled. The sysadmin user cannot be deleted. Admin users cannot delete security officers. Only security officers can delete, enable, and disable other security officers. Procedure
1. Select Administration > Access > Local Users. The Local Users view appears.
2. Click one or more user names from the list. 3. Click Delete to delete the user accounts. The Delete User dialog box appears.
4. Click OK and Close.
Enabling and disabling local users Admin users can enable or disable all users except the sysadmin user and users with the security role. The sysadmin user cannot be disabled. Only Security officers can enable or disable other security officers. Procedure
1. Select Administration > Access > Local Users. The Local Users view appears.
Local user account management 91 Managing Data Domain Systems
2. Click one or more user names from the list. 3. Click either Enable or Disable to enable or disable user accounts. The Enable or Disable User dialog box appears.
4. Click OK and Close.
Enabling security authorization You can use the Data Domain system command-line interface (CLI) to enable and disable the security authorization policy. For information on the commands used in this procedure, see the EMC Data Domain Operating System Command Reference Guide.
The Retention Lock Compliance license must be installed. You are not permitted to disable the authorization policy on Retention Lock Compliance systems.
Procedure 1. Log into the CLI using a security officer username and password. 2. To enable the security officer authorization policy, enter: # authorization policy set security-officer enabled
Changing user passwords After you create a user, you can use DD System Manager to change the user's password. Individual users can also change their own passwords. Procedure
1. Click Administration > Access > Local Users. The Local Users view appears.
2. Click a user name from the list. 3. Click Change Password to change the user password. The Change Password dialog box appears. If prompted, enter your old password.
4. Enter the new password into the New Password box. 5. Enter the new password again into Verify New Password box. 6. Click OK.
Modifying the password policy and login controls The password policy and login controls define login requirements for all users. Administrators can specify how often a password must be changed, what is required to create a valid password, and how the system responds to invalid login attempts. Procedure
1. Select Administration > Access. 2. Select More Tasks > Change Login Options. The Change Login Options dialog appears.
92 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
3. Specify the new configuration in the boxes for each option. To select the default value, click Default next to the appropriate option. 4. Click OK to save the password settings. Change Login Options dialog Use this dialog to set the password policy and specify the maximum login attempts and lockout period.
Table 32 Change Login Options dialog controls
Minimum Days Between The minimum number of days between password changes that you Change allow a user. This value must be less than the Maximum Days Between Change value minus the Warn Days Before Expire value. The default setting is 0. Maximum Days Between The maximum number of days between password changes that you Change allow a user. The minimum value is 1. The default value is 90. Warn Days Before Expire The number of days to warn the users before their password expires. This value must be less than the Maximum Days Between Change value minus the Minimum Days Between Change value. The default setting is 7. Disable Days After Expire The system disables a user account after password expiration according to the number of days specified with this option. Valid entries are never or number greater than or equal to 0. The default setting is never. Minimum Length of The minimum password length required. Default is 6. Password Minimum Number of The minimum number of character classes required for a user Character Classes password. Default is 1. Character classes include:
l Lowercase letters (a-z)
l Uppercase letters (A-Z)
l Special Characters ($, %, #, +, and so on)
Lowercase Character Enable or disable the requirement for at least one lowercase Requirement character. The default setting is disabled. Uppercase Character Enable or disable the requirement for at least one uppercase Requirement character. The default setting is disabled. One Digit Requirement Enable or disable the requirement for at least one numerical character. The default setting is disabled. Special Character Enable or disable the requirement for at least one special character. Requirement The default setting is disabled. Max Consecutive Enable or disable the requirement for a maximum of three repeated Character Requirement characters. The default setting is disabled. Prevent use of Last N Specify the number of remembered passwords. The range is 0 to 24, Passwords and the default settings is 1.
Local user account management 93 Managing Data Domain Systems
Table 32 Change Login Options dialog controls (continued)
If this setting is reduced, the remembered password list remains unchanged until the next time the password is changed. For example, if this setting is changed from 4 to 3, the last four passwords are remembered until the next time the password is changed.
Maximum login attempts Specifies the maximum number of login attempts before a mandatory lock is applied to a user account. This limit applies to all user accounts, including sysadmin. A locked user cannot log in while the account is locked. The range is 4 to 20, and the default value is 4. Unlock timeout (seconds) Specifies how long a user account is locked after the maximum number of login attempts. When the configured unlock timeout is reached, a user can attempt login. The range is 120 to 3600 seconds, and the default period is120 seconds.
Directory user and group management You can use DD System manager to manage access to the system for users and groups in Windows Active Directory, Windows Workgroup, and NIS. Kerberos authentication is an option for CIFS and NFS clients.
Viewing Active Directory and Kerberos information The Active Directory Kerberos configuration determines the methods CIFS and NFS clients use to authenticate. The Active Directory/Kerberos Authentication panel displays this configuration. Procedure
1. Select Administration > Access > Authentication. 2. Expand the Active Directory/Kerberos Authentication panel.
Table 33 Active Directory/ Kerberos Authentication label descriptions
Mode The type of authentication mode. In Windows/Active Directory mode, CIFS clients use Active Directory and Kerberos authentication, and NFS clients use Kerberos authentication. In Unix mode, CIFS clients use Workgroup authentication (without Kerberos), and NFS clients use Kerberos authentication. In Disabled mode, Kerberos authentication is disabled and CIFS clients use Workgroup authentication. Realm The realm name of the Workgroup or Active Directory. DDNS Whether or not the Dynamic Domain Name System is enabled. Domain Controllers The name of the domain controller for the Workgroup or Active Directory.
94 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 33 Active Directory/ Kerberos Authentication label descriptions (continued)
Organizational Unit The name of the organizations unit for the Workgroup or Active Directory. CIFS Server Name The name of the CIFS server in use (Windows mode only). WINS Server The name of the WINS server in use (Windows mode only). Short Domain Name An abbreviated name for the domain. NTP Enabled/Disabled (UNIX mode only) NIS Enabled/Disabled (UNIX mode only) Key Distribution Centers Hostname(s) or IP(s) of KDC in use (UNIX mode only) Active Directory Enabled/Disabled: Click to Enable or disable administrative access Administrative Access for Active Directory (Windows) groups.
Table 34 Active Directory administrative groups and roles
Windows Group The name of the Windows group. Management Role The role of the group (admin, user, and so on)
Configuring Active Directory and Kerberos authentication Configuring Active Directory authentication makes the Data Domain system part of a Windows Active Directory realm. CIFS clients and NFS clients use Kerberos authentication. Procedure 1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the Active Directory/Kerberos Authentication panel. 3. Click Configure. next to Mode to start the configuration wizard. The Active Directory/Kerberos Authentication dialog appears. 4. Select Windows/Active Directory and click Next. 5. Enter the full realm name for the system (for example: domain1.local), the user name, and password for the Data Domain system. Then click Next.
Use the complete realm name. Ensure that the user is assigned sufficient privileges to join the system to the domain. The user name and password must be compatible with Microsoft requirements for the Active Directory domain. This user must also be assigned permission to create accounts in this domain.
6. Select the default CIFS server name, or select Manual and enter a CIFS server name. 7. To select domain controllers, select Automatically assign, or select Manual and enter up to three domain controller names.
Directory user and group management 95 Managing Data Domain Systems
You can enter fully qualified domain names, hostnames, or IP (IPv4 or IPv6) addresses. 8. To select an organizational unit, select Use default Computers, or select Manual and enter an organization unit name.
The account is moved to the new organizational unit.
9. Click Next. The Summary page for the configuration appears. 10. Click Finish. The system displays the configuration information in the Authentication view. 11. To enable administrative access, click Enable to the right of Active Directory Administrative Access. Authentication mode selections The authentication mode selection determines how CIFS and NFS clients authenticate using supported combinations of Active Directory, Workgroup, and Kerberos authentication. DD OS supports the following authentication options.
l Disabled: Kerberos authentication is disabled for CIFS and NFS clients. CIFS clients use Workgroup authentication. l Windows/Active Directory: Kerberos authentication is enabled for CIFS and NFS clients. CIFS clients use Active Directory authentication. l Unix: Kerberos authentication is enabled for only NFS clients. CIFS clients use Workgroup authentication.
Managing administrative groups for Active Directory You can use the Active Directory/Kerberos Authentication panel to create, modify, and delete Active Directory (Windows) groups and assign management roles (admin, backup- operator, and so on) to those groups. To prepare for managing groups, select Administration > Access > Authentication , expand the Active Directory/Kerberos Authentication panel, and click the Active Directory Administrative Access Enable button. Creating administrative groups for Active Directory Create an administrative group when you want to assign a management role to all the users configured in an Active Directory group. Before you begin Enable Active Directory Administrative Access on the Active Directory/Kerberos Authentication panel in the Administration > Access > Authentication page. Procedure 1. Click Create. 2. Enter the domain and group name separated by a backslash. For example: domainname\groupname. 3. Select the management role for the group from the drop-down menu.
96 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
4. Click OK. Modifying administrative groups for Active Directory Modify an administrative group when you want to change the administrative group name or management role configured for an Active Directory group. Before you begin Enable Active Directory Administrative Access on the Active Directory/Kerberos Authentication panel in the Administration > Access > Authentication page. Procedure 1. Select a group to modify under the Active Directory Administrative Access heading. 2. Click Modify. 3. Modify the domain and group name. These names are separated by a backslash. For example: domainname\groupname. 4. Modify the management role for the group by selecting a different role from the drop- down menu. Deleting administrative groups for Active Directory Delete an administrative group when you want to terminate system access for all the users configured in an Active Directory group. Before you begin Enable Active Directory Administrative Access on the Active Directory/Kerberos Authentication panel in the Administration > Access > Authentication page. Procedure 1. Select a group to delete under the Active Directory Administrative Access heading. 2. Click Delete.
Configuring UNIX Kerberos authentication Configuring UNIX Kerberos authentication enables NFS clients to use Kerberos authentication. CIFS clients use Workgroup authentication. Before you begin NIS must be running for UNIX-mode Kerberos authentication to function. For instructions about enabling Kerberos, see the section regarding enabling NIS services. Configuring Kerberos for UNIX enables NFS clients to use Kerberos authentication. CIFS clients use Workgroup authentication. Procedure
1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the Active Directory/Kerberos Authentication panel. 3. Click Configure. next to Mode to start the configuration wizard. The Active Directory/Kerberos Authentication dialog appears. 4. Select Unix and click Next. 5. Enter the realm name (for example: domain1.local), and up to three host names or IP addresses (IPv4 or IPv6) for key distribution centers (KDCs).
Directory user and group management 97 Managing Data Domain Systems
6. Optionally, click Browse to upload a keytab file, and click Next. The Summary page for the configuration appears.
Keytab files are generated on the authentication servers (KDCs) and contain a shared secret between the KDC server and the DDR.
A keytab file must be uploaded and imported for Kerberos authentication to operate correctly.
7. Click Finish. The system displays the configuration information in the Active Directory/Kerberos Authentication panel.
Disabling Kerberos authentication Disabling Kerberos authentication prevents CIFS and NFS clients from using Kerberos authentication. CIFS clients use Workgroup authentication. Procedure
1. Select Administration > Access Management > Authentication. The Authentication view appears.
2. Expand the Active Directory/Kerberos Authentication panel. 3. Click Configure. next to Mode to start the configuration wizard. The Active Directory/Kerberos Authentication dialog appears. 4. Select Disabled and click Next. The system displays a summary page with changes appearing in bold text. 5. Click Finish. The system displays Disabled next to Mode in the Active Directory/Kerberos Authentication panel.
Viewing Workgroup authentication information Use the Workgroup Authentication panel to view Workgroup configuration information. Procedure
1. Select Administration > Access > Authentication. 2. Expand the Workgroup Authentication panel.
98 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 35 Workgroup Authentication label descriptions
Mode The type of authentication mode (Workgroup or Active Directory). Workgroup name The specified workgroup CIFS Server Name The name of the CIFS server in use. WINS Server The name of the WINS server in use.
Configuring workgroup authentication parameters Workgroup authentication parameters allow you to configure a Workgroup name and CIFS server name. Procedure
1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the Workgroup Authentication panel. 3. Click Configure. The Workgroup Authentication dialog appears.
4. For Workgroup Name, select Manual and enter a workgroup name to join, or use the default. The Workgroup mode joins a Data Domain system to a workgroup domain.
5. For CIFS Server Name, select Manual and enter a server name (the DDR), or use the default. 6. Click OK.
Viewing NIS authentication information The NIS Authentication panel displays the NIS configuration parameters and whether NIS authentication is enabled or disabled. Procedure
1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the NIS Authentication panel. Results
Table 36 NIS Authentication panel items
NIS Status Enabled or Disabled. Domain Name The name of the domain for this service. Server Authentication server(s).
Directory user and group management 99 Managing Data Domain Systems
Table 36 NIS Authentication panel items (continued)
NIS Group The name of the NIS group. Management Role The role of the group (admin, user, and so on).
Enabling and disabling NIS authentication Use the NIS Authentication panel to enable and disable NIS authentication. Procedure
1. Select Maintenance > Access > Authentication. The Authentication view appears.
2. Expand the NIS Authentication panel. 3. Click Enable next to NIS Status to enable or Disable to disable NIS Authentication. The Enable or Disable NIS dialog box appears.
Configuring the NIS domain name Use the NIS Authentication panel to configure the NIS domain name. Procedure
1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the NIS Authentication panel. 3. Click Edit next to Domain Name to edit the NIS domain name. The Configure NIS Domain Name dialog box appears.
4. Enter the domain name in the Domain Name box. 5. Click OK.
Specifying NIS authentication servers Use the NIS Authentication panel to specify NIS authentication servers. Procedure
1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the NIS Authentication panel. 3. Below Domain Name, select one of the following:
l Obtain NIS Servers from DHCP The system automatically obtains NIS servers using DHCP l Manually Configure Use the following procedures to manually configure NIS servers.
100 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
l To add an authentication server, click Add (+) in the server table, enter the server name, and click OK. l To modify an authentication server, select the authentication server name and click the edit icon (pencil). Change the server name, and click OK. l To remove an authentication server name, select a server, click the X icon, and click OK. 4. Click OK.
Configuring NIS groups Use the NIS Authentication panel to configure NIS groups. Procedure
1. Select Administration > Access > Authentication. The Authentication view appears.
2. Expand the NIS Authentication panel. 3. Configure the NIS groups in the NIS Group table.
l To add a NIS group, click Add (+), enter the NIS group name and role, and click Validate. Click OK to exit the add NIS group dialog box. Click OK again to exit the Configure Allowed NIS Groups dialog box. l To modify an NIS group, select the checkbox of the NIS group name in the NIS group list and click Edit (pencil). Change the NIS group name, and click OK. l To remove an NIS group name, select the NIS group in the list and click Delete X.
Configuring mail server settings The Mail Server tab allows you to specify the mail server to which DD OS sends email reports. Procedure
1. Select Administration > Settings > Mail Server. 2. Select More Tasks > Set Mail Server. The Set Mail Server dialog box appears.
3. Enter the name of the mail server in the Mail Server box. 4. Click OK.
Managing time and date settings The Time and Date Settings tab allows you to view and configure the system time and date or configure the Network Time Protocol to set the time and date. Procedure
1. To view the current time and date configuration, select Administration > Settings > Time and Date Settings.
Configuring mail server settings 101 Managing Data Domain Systems
The Time and Date Settings page presents the current system date and time, shows whether NTP is enabled or not, and lists the IP addresses or hostnames of configured NTP servers.
2. To change the configuration, select More Tasks > Configure Time Settings. The Configure Time Settings dialog appears.
3. In the Time Zone dropdown list, select the time zone where the Data Domain system resides. 4. To manually set the time and date, select None, type the date in the Date box, and select the time in the Time dropdown lists. 5. To use NTP to synchronize the time, select NTP and set how the NTP server is accessed.
l To use DHCP to automatically select a server, select Obtain NTP Servers using DHCP. l To configure an NTP server IP address, select Manually Configure, add the IP address of the server, and click OK.
Using time synchronization from an Active Directory domain controller might cause excessive time changes on the system if both NTP and the domain controller are modifying the time.
6. Click OK. 7. If you changed the time zone, you must reboot the system.
a. Select Maintenance > System. b. From the More Tasks menu, select Reboot System. c. Click OK to confirm.
Managing system properties The System Properties tab allows you to view and configure system properties that identify the managed system location, administrator email address, and host name. Procedure
1. To view the current configuration, select Administration > Settings > System Properties. The System Properties tab displays the system location, the administrator email address, and the administrator hostname.
2. To change the configuration, select More Tasks > Set System Properties. The Set System Properties dialog box appears.
3. In the Location box, enter information about where the Data Domain system is located. 4. In the Admin Email box, enter the email address of the system administrator. 5. In the Admin Host box, enter the name of the administration server.
102 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
SNMP management The Simple Network Management Protocol (SNMP) is a standard protocol for exchanging network management information, and is a part of the Transmission Control Protocol/ Internet Protocol (TCP/IP) protocol suite. SNMP provides a tool for network administrators to manage and monitor network-attached devices, such as Data Domain systems, for conditions that warrant administrator attention. To monitor Data Domain systems using SNMP, you will need to install the Data Domain MIB in your SNMP Management system. DD OS also supports the standard MIB-II so you can also query MIB-II statistics for general data such as network statistics. For full coverage of available data you should utilize both the Data Domain MIB and the standard MIB-II MIB. The Data Domain system SNMP agent accepts queries for Data Domain-specific information from management systems using SNMP v1, v2c, and v3. SNMP V3 provides a greater degree of security than v2c and v1 by replacing cleartext community strings (used for authentication) with user-based authentication using either MD5 or SHA1. Also, SNMP v3 user authentication packets can be encrypted and their integrity verified with either DES or AES. Data Domain systems can send SNMP traps (which are alert messages) using SNMP v2c and SNMP v3. Because SNMP v1 traps are not supported, EMC recommends using SNMP v2c or v3. The default port that is open when SNMP is enabled is port 161. Traps are sent out through port 162.
l The EMC Data Domain Operating System Initial Configuration Guide describes how to set up the Data Domain system to use SNMP monitoring. l The EMC Data Domain Operating System MIB Quick Reference describes the full set of MIB parameters included in the Data Domain MIB branch.
Viewing SNMP status and configuration The SNMP tab displays the current SNMP status and configuration. Procedure
1. Select Administration > Settings > SNMP. The SNMP view shows the SNMP status, SNMP properties, SNMP V3 configuration, and SNMP V2C configuration.
SNMP tab labels The SNMP tab labels identify the overall SNMP status, SNMP property values, and the configurations for SNMPv3 and SNMPv2. Status The Status area displays the operational status of the SNMP agent on the system, which is either Enabled or Disabled.
SNMP management 103 Managing Data Domain Systems
SNMP Properties Table 37 SNMP Properties descriptions
SNMP System Location The location of the Data Domain system being monitored. SNMP System Contact The person designated as the person to contact for the Data Domain system administration.
SNMP V3 Configuration Table 38 SNMP Users column descriptions
Name The name of the user on the SNMP manager with access to the agent for the Data Domain system.
Access The access permissions for the SNMP user, which can be Read- only or Read-write. Authentication Protocols The Authentication Protocol used to validate the SNMP user, which can be MD5, SHA1, or None. Privacy Protocol The encryption protocol used during the SNMP user authentication, which can be AES, DES, or None.
Table 39 Trap Hosts column descriptions
Host The IP address or domain name of the SNMP management host. Port The port used for SNMP trap communication with the host. For example, 162 is the default. User The user on the trap host authenticated to access the Data Domain SNMP information.
SNMP V2C Configuration Table 40 Communities column descriptions
Community The name of the community. For example, public, private, or localCommunity. Access The access permission assigned, which can be Read-only or Read- write. Hosts The hosts in this community.
104 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 41 Trap Hosts column descriptions
Host The systems designated to receive SNMP traps generated by the Data Domain system. If this parameter is set, systems receive alert messages, even if the SNMP agent is disabled. Port The port used for SNMP trap communication with the host. For example, 162 is the default. Community The name of the community. For example, public, private, or localCommunity.
Enabling and disabling SNMP Use the SNMP tab to enable of disable SNMP. Procedure
1. Select Administration > Settings > SNMP. 2. In the Status area, click Enable or Disable.
Downloading the SNMP MIB Use the SNMP tab to download the SNMP MIB. Procedure
1. Select Administration > Settings > SNMP. 2. Click Download MIB file. 3. In the Opening DATA_DOMAIN.mib dialog box, select Open. 4. Click Browse and select a browser to view the MIB in a browser window.
If using the Microsoft Internet Explorer browser, enable Automatic prompting for file download.
5. Save the MIB or exit the browser.
Configuring SNMP properties Use the SNMP tab to configure the text entries for system location and system contact. Procedure
1. Select Administration > Settings > SNMP. 2. In the SNMP Properties area, click Configure. The SNMP Configuration dialog box appears.
3. In the text fields, add an SNMP system location (a description of where the Data Domain system is located) and/or an SNMP system contact (for example, the email address of the system administrator for the Data Domain system). 4. Click OK.
Enabling and disabling SNMP 105 Managing Data Domain Systems
SNMP V3 user management Use the SNMP tab to create, modify, and delete SNMPv3 users and trap hosts.
Creating SNMP V3 users When you create SNMPv3 users, you define a username, specify either read-only or read- write access, and select an authentication protocol. Procedure
1. Select Administration > Settings > SNMP. 2. In the SNMP Users area, click Create. The Create SNMP User dialog box appears.
3. In the Name text field, enter the name of the user for whom you want to grant access to the Data Domain system agent. The name must be a minimum of eight characters. 4. Select either read-only or read-write access for this user. 5. To authenticate the user, select Authentication. a. Select either the MD5 or the SHA1 protocol. b. Enter the authentication key in the Key text field. c. To provide encryption to the authentication session, select Privacy. d. Select either the AES or the DES protocol. e. Enter the encryption key in the Key text field. 6. Click OK. The newly added user account appears in the SNMP Users table.
Modifying SNMP V3 users You can modify the access level (read-only or read-write) and authentication protocol for existing SNMPv3 users. Procedure
1. Select Administration > Settings > SNMP. 2. In the SNMP Users area, select a checkbox for the user and click Modify. The Modify SNMP User dialog box appears. Add or change any of the following settings.
3. Select either read-only or read-write access for this user. 4. To authenticate the user, select Authentication. a. Select either the MD5 or the SHA1 protocol. b. Enter the authentication key in the Key text field. c. To provide encryption to the authentication session, select Privacy. d. Select either the AES or the DES protocol. e. Enter the encryption key in the Key text field. 5. Click OK.
106 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
The new settings for this user account appear in the SNMP Users table.
Removing SNMP V3 users Use the SNMP tab to delete existing SNMPv3 users. Procedure
1. Select Administration > Settings > SNMP. 2. In the SNMP Users area, select a checkbox for the user and click Delete. The Delete SNMP User dialog box appears.
If the Delete button is disabled, the selected user is being used by one or more trap hosts. Delete the trap hosts and then delete the user.
3. Verify the user name to be deleted and click OK. 4. In the Delete SNMP User Status dialog box, click Close. The user account is removed from the SNMP Users table.
SNMP V2C community management Define SNMP v2c communities (which serve as passwords) to control management system access to the Data Domain system. To restrict access to specific hosts that use the specified community, assign the hosts to the community.
The SNMP V2c Community string is a sent in cleartext and is very easy to intercept. If this occurs, the interceptor can retrieve information from devices on your network, modify their configuration, and possibly shut them down. SNMP V3 provides authentication and encryption features to prevent interception.
SNMP community definitions do not enable the transmission of SNMP traps to a management station. You must define trap hosts to enable trap submission to management stations.
Creating SNMP V2C communities Create communities to restrict access to the DDR system or for use in sending traps to a trap host. You must create a community and assign it to a host before you can select that community for use with the trap host. Procedure
1. Select Administration > Settings > SNMP. 2. In the Communities area, click Create. The Create SNMP V2C Community dialog box appears.
3. In the Community box, enter the name of a community for whom you want to grant access to the Data Domain system agent.
SNMP V2C community management 107 Managing Data Domain Systems
4. Select either read-only or read-write access for this community. 5. If you want to associate the community to one or more hosts, add the hosts as follows: a. Click + to add a host. The Host dialog box appears. b. In the Host text field, enter the IP address or domain name of the host. c. Click OK. The Host is added to the host list. 6. Click OK. The new community entry appears in the Communities table and lists the selected hosts.
Modifying SNMP V2C Communities Procedure
1. Select Administration > Settings > SNMP. 2. In the Communities area, select the checkbox for the community and click Modify. The Modify SNMP V2C Community dialog box appears.
3. To change the access mode for this community, select either read-only or read-write access.
The Access buttons for the selected community are disabled when a trap host on the same system is configured as part of that community. To modify the access setting, delete the trap host and add it back after the community is modified.
4. To add one or more hosts to this community, do the following: a. Click + to add a host. The Host dialog box appears. b. In the Host text field, enter the IP address or domain name of the host. c. Click OK. The Host is added to the host list. 5. To delete one or more hosts from the host list, do the following:
DD System Manager does not allow you to delete a host when a trap host on the same system is configured as part of that community. To delete a trap host from a community, delete the trap host and add it back after the community is modified.
108 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
The Access buttons for the selected community are not disabled when the trap host uses an IPv6 address and the system is managed by an earlier DD OS version that does not support IPv6. EMC recommends that you always select a management system that uses the same or a newer DD OS version than the systems it manages.
a. Select the checkbox for each host or click the Host check box in the table head to select all listed hosts. b. Click the delete button (X). 6. To edit a host name, do the following: a. Select the checkbox for the host. b. Click the edit button (pencil icon). c. Edit the host name. d. Click OK. 7. Click OK. The modified community entry appears in the Communities table.
Deleting SNMP V2C communities Use the SNMP tab to delete existing SNMPv2 communities. Procedure
1. Select Administration > Settings > SNMP. 2. In the Communities area, select a checkbox for the community and click Delete. The Delete SNMP V2C Communities dialog box appears.
If the Delete button is disabled, the selected community is being used by one or more trap hosts. Delete the trap hosts and then delete the community.
3. Verify the community name to be deleted and click OK. 4. In the Delete SNMP V2C Communities Status dialog box, click Close. The community entry is removed from the Communities table.
SNMP trap host management Trap host definitions enable Data Domain systems to send alert messages in SNMP trap messages to an SNMP management station.
Creating SNMP V3 and V2C trap hosts Trap host definitions identify remote hosts that receive SNMP trap messages from the system. Before you begin If you plan to assign an existing SNMP v2c community to a trap host, you must first use the Communities area to assign the trap host to the community.
SNMP trap host management 109 Managing Data Domain Systems
1. Select Administration > Settings > SNMP. 2. In the SNMP V3 Trap Hosts or SNMP V2C Trap Hosts area, click Create. The Create SNMP [V3 or V2C] Trap Hosts dialog appears.
3. In the Host box, enter the IP address or domain name of the SNMP Host to receive traps. 4. In the Port box, enter the port number for sending traps (port 162 is a common port). 5. Select the user (SNMP V3) or the community (SNMP V2C) from the drop-down menu.
The Community list displays only those communities to which the trap host is already assigned.
6. To create a new community, do the following: a. Select Create New Community in the Community drop-down menu. b. Enter the name for the new community in the Community box. c. Select the Access type. d. Click the add (+) button. e. Enter the trap host name. f. Click OK. g. Click OK. 7. Click OK.
Modifying SNMP V3 and V2C trap hosts You can modify the port number and community selection for existing trap host configurations. Procedure
1. Select Administration > Settings > SNMP. 2. In the SNMP V3 Trap Hosts or SNMP V2C Trap Hosts area, select a Trap Host entry, and click Modify. The Modify SNMP [V3 or V2C] Trap Hosts dialog box appears.
3. To modify the port number, enter a new port number in the Port box (port 162 is a common port). 4. Select the user (SNMP V3) or the community (SNMP V2C) from the drop-down menu.
The Community list displays only those communities to which the trap host is already assigned.
5. To create a new community, do the following: a. Select Create New Community in the Community drop-down menu.
110 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
b. Enter the name for the new community in the Community box. c. Select the Access type. d. Click the add (+) button. e. Enter the trap host name. f. Click OK. g. Click OK. 6. Click OK.
Removing SNMP V3 and V2C trap hosts Use the SNMP tab to delete existing trap host configurations. Procedure
1. Select Administration > Settings > SNMP. 2. In the Trap Hosts area (either for V3 or V2C, select a checkbox for the trap host and click Delete. The Delete SNMP [V3 or V2C] Trap Hosts dialog box appears.
3. Verify the host name to be deleted and click OK. 4. In the Delete SNMP [V3 or V2C] Trap Hosts Status dialog box, click Close. The trap host entry is removed from the Trap Hosts table.
Autosupport report management The Autosupport feature generates a report called an ASUP. The ASUP shows system identification information, consolidated output from a number of Data Domain system commands, and entries from various log files. Extensive and detailed internal statistics appear at the end of the report. This report is designed to aid EMC Data Domain Support in debugging system problems. An ASUP is generated every time the file system is started, which is usually once per day. However, the file system can be started more than once in a day. You can configure email addresses to receive the daily ASUP reports, and you can enable or disable sending of these reports to EMC Data Domain. The default time for sending the daily ASUP is 06.00 a.m, and it is configurable. When sending ASUPs to EMC, you have the option to select the legacy unsecure method or the ConnectEMC method, which encrypts the information before transmission.
HA system autosupport and support bundle manageability Configuration is done on the active node and mirrored to the standby node; therefore, the same configuration is on both nodes, but there is not a consolidated ASUP and support bundle. Auto-support and support bundle on active node also includes filesystem, replication, protocol, and full HA information in addition to local node information, and auto-support and support bundle on the standby node only have local node information plus some HA information (configuration and status), no filesystem/replication/protocol information. The auto-supports and support bundles from both the nodes will be needed to debug issues related to HA system status (filesystem, replication, protocols, and HA configuration).
Autosupport report management 111 Managing Data Domain Systems
Enabling and disabling autosupport reporting to EMC You can enable or disable autosupport reporting to EMC without affecting whether or not alerts are sent to EMC. Procedure
1. To view the autosupport reporting status, select Maintenance > Support > Autosupport. The autosupport reporting status is highlighted next to the Scheduled auto support label in the EMC Support area. Depending on the current configuration, either an Enable or a Disable button appears in the Scheduled auto support row. 2. To enable alert reporting to EMC, click Enable in the Scheduled auto support row. 3. To disable alert reporting to EMC, click Disable in the Scheduled auto support row.
Reviewing generated autosupport reports Review autosupport reports to view system statistics and configuration information captured in the past. The system stores a maximum of 14 autosupport reports. Procedure
1. Select Maintenance > Support > Autosupport. The Autosupport Reports page shows the autosupport report file name and file size, and the date the report was generated. Reports are automatically named. The most current report is autosupport, the previous day is autosupport.1, and the number increments as the reports move back in time.
2. Click the file name link to view the report using a text editor. If doing so is required by your browser, download the file first.
Configuring the autosupport mailing list Autosupport mailing list subscribers receive autosupport messages through email. Use the Autosupport tab to add, modify, and delete subscribers. Autosupport emails are sent through the configured mail server to all subscribers in the autosupport email list. After you configure the mail server and autosupport email list, it is a good practice to test the setup to ensure that autosupport messages reach the intended destinations. Procedure
1. Select Maintenance > Support > Autosupport. 2. Click Configure. The Configure Autosupport Subscribers dialog box appears.
3. To add a subscriber, do the following. a. Click Add (+). The Email dialog box appears. b. Enter the recipients email address in the Email box. c. Click OK. 4. To delete a subscriber, do the following.
112 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
a. In the Configure Autosupport Subscribers dialog box, select the subscriber to delete. b. Click Delete (X). 5. To modify a subscriber email address, do the following. a. In the Configure Autosupport Subscribers dialog box, select the subscriber name to edit. b. Click Modify (pencil icon). The Email dialog box appears. c. Modify the email address as needed. d. Click OK. 6. Click OK to close the Configure Autosupport Subscribers dialog box. The revised autosupport email list appears in the Autosupport Mailing List area.
Support bundle management A support bundle is a file that contains system configuration and operation information. It is a good practice to generate a support bundle before a software upgrade or a system topology change (such as a controller upgrade). EMC Data Domain Support often requests a support bundle when providing assistance.
Generating a support bundle When troubleshooting problems, EMC Data Domain Customer Support may ask for a support bundle, which is a tar-g-zipped selection of log files with a README file that includes identifying autosupport headers. Procedure
1. Select Maintenance > Support > Support Bundles. 2. Click Generate Support Bundle.
The system supports a maximum of 10 support bundles. If you attempt to generate an 11th support bundle, DD System Manager displays a warning that prompts you to click OK or Cancel. If you click OK to continue, the oldest support bundle, which is listed in the warning message, is deleted. You can also delete support bundles using the CLI command support bundle delete.
Also, if you generate a support bundle on a upgraded system that contains a support bundle named using the old format, support-bundle.tar.gz, that file is renamed to use the newer name format.
3. Email the file to customer support at [email protected].
If the bundle is too large to be emailed, use the EMC support site to upload the bundle. (Go to https://support.emc.com.)
Support bundle management 113 Managing Data Domain Systems
Viewing the support bundles list Use the Support Bundles tab to view the support bundle files on the system. Procedure 1. Select Maintenance > Support > Support Bundles. The Support Bundles list appears. Listed are the support bundle file name, file size, and date the bundle was generated. Bundles are automatically named hostname-support-bundle- datestamp.tar.gz. An example filename is localhost-support- bundle-1127103633.tar.gz, which indicates that the support bundle was created on the localhost system on November 27th at 10:36:33. 2. Click the file name link and select a gz/tar decompression tool to view the ASCII contents of the bundle.
Alert notification management The alert feature generates event and summary reports that can be distributed to configurable email lists and to EMC Data Domain. Event reports are sent immediately and provide detailed information on a system event. The distribution lists for event alerts are called notification groups. You can configure a notification group to include one or more email addresses, and you can configure the types and severity level of the event reports sent to those addresses. For example, you might configure one notification group for individuals who need to know about critical events and another group for those who monitor less critical events. Another option is to configure groups for different technologies. For example, you might configure one notification group to receive emails about all network events and another group to receive messages about storage issues. Summary reports are sent daily and provide a summary of the events that occurred during the last 24 hours. Summary reports do not include all the information provided in event reports. The default generation time for the daily report is 08.00 a.m, and it can be changed. Summary reports are sent using a dedicated email list that is separate from the event notification groups. You can enable or disable alert distribution to EMC Data Domain. When sending reports to EMC, you have the option to select the legacy unsecure method or the ConnectEMC method, which encrypts the information before transmission.
HA system alert notification management The alert feature on an HA system generates event and summary report like a non-HA system but how the HA system manages these alerts is different due to the two node system set-up. Initial alert configuration is completed on the active node and mirrored to the stand-by (i.e., same configuration on both nodes). Local and AM-Alerts are emailed according to the notification settings and include information indicating they are from an HA system and from which node, the active or standby, that generated the alerts. If there are active alerts on the file system, replication, or protocols when a failover occurs, these active alerts continue to show on the new active node after failover if the alert conditions have not cleared up. Historical alerts on the filesystem, replication, and protocols stay with the node where they originated rather than failing over together with the filesystem on a failover. This
114 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
means the CLIs on the active node will not present a complete/continuous view of historical alerts for filesystem, replication, and protocols During a failover, local historical alerts stay with the node from which they were generated; however, the historical alerts for the filesystem, replication, and protocols (generally called "logical alerts") fail over together with the filesystem.
The Health > High Availability panel displays only alerts that are HA-related. Those alerts can be filtered by major HA component, such as HA Manager, Node, Interconnect, Storage, and SAS connection.
Viewing the notification group list A notification group defines a set of alert types (classes) and a group of email addresses (for subscribers). Whenever the system generates an alert type selected in a notification list, that alert is sent to the list subscribers. Procedure
1. Select Health > Alerts > Notification. 2. To limit (filter) the entries in the Group Name list, type a group name in the Group Name box or a subscriber email in the Alert Email box, and click Update.
Click Reset to display all configured groups.
3. To display detailed information for a group, select the group in the Group Name list.
Notification tab The Notification tab allows you to configure groups of email address that receive system alerts for the alert types and severity levels you select.
Table 42 Group Name list, column label descriptions
Group Name The configured name for the group. Classes The number of alert classes that are reported to the group. Subscribers The number of subscribers who are configured to receive notifications through email.
Table 43 Detailed Information, label descriptions
Class A service or subsystem that can forward alerts. The listed classes are those for which the notification group receives alerts. Severity The severity level that triggers an email to the notification group. All alerts at the specified severity level and above are sent to the notification group.
Viewing the notification group list 115 Managing Data Domain Systems
Table 43 Detailed Information, label descriptions (continued)
Subscribers The subscribers area displays a list of all email addresses configured for the notification group.
Table 44 Notification tab controls
Add button Click the Add button to begin creating a notification group.
Class Attributes Configure button Click this Configure button to change the classes and severity levels that generate alerts for the selected notification group.
Delete button Click the Delete button to delete the selected notification group.
Filter By: Alert Email box Enter text in this box to limit the group name list entries to groups that include an email address that contains the specified text.
Filter By: Group Name box Enter text in this box to limit the group name list entries to group names that contain the specified text.
Modify button Click the Modify button to modify the configuration for the selected notification group.
Reset button Click this button to remove any entries in the Filter By boxes and display all group names.
Subscribers Configure button Click this Configure button to change the email list for the selected notification group.
Update button Click this button to update the group name list after you enter text in a filter box.
Creating a notification group Use the Notification tab to add notification groups and select the severity level for each group. Procedure 1. Select Health > Alerts > Notification. 2. Click Add. The Add Group dialog box appears.
3. Type the group name in the Group Name box. 4. Select the checkbox of one or more alert classes of which to be notified. 5. To change the default severity level (Warning) for a class, select another level in the associated list box.
116 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
The severity levels are listed in ascending severity level. Emergency is the highest severity level. 6. Click OK.
Managing the subscriber list for a group Use the Notification tab to add, modify, or delete email addresses from a notification group subscriber list. Procedure
1. Select Health > Alerts > Notification. 2. Select the checkbox of a group in the Notifications group list, and do one of the following.
l Click Modify and select Subscribers. l Click Configure in the Subscribers list. 3. To add a subscriber to the group, do the following. a. Click the + icon. The Email Address dialog box appears.
b. Enter the email address of a subscriber. c. Click OK. 4. To modify an email address, do the following. a. Click the checkbox of the email address in the Subscriber Email list. b. Click the pencil icon. c. Edit the email address in the Email Address dialog box. d. Click OK. 5. To delete an email address, click the checkbox of the email address in the Subscriber Email list and click the X icon. 6. Click Finish or OK.
Modifying a notification group Use the Notification table to modify the attribute classes in an existing group. Procedure
1. Select Health > Alerts > Notification. 2. Select the checkbox of the group to modify in the group list. 3. To modify the class attributes for a group, do the following. a. Click Configure in the Class Attributes area. The Edit Group dialog box appears. b. Select (or clear) the checkbox of one or more class attributes. c. To change the severity level for a class attribute, select a level from the corresponding list box. d. Click OK.
Managing the subscriber list for a group 117 Managing Data Domain Systems
4. To modify the subscriber list for a group, do the following. a. Click Configure in the Subscribers area. The Edit Subscribers dialog box appears. b. To delete subscribers from the group list, select the checkboxes of subscribers to delete and click the Delete icon (X). c. To add a subscriber, click the Add icon (+), type a subscriber email address, and click OK. d. Click OK. 5. Click OK.
Deleting a notification group Use the Notification tab to delete one or more existing notification groups. Procedure 1. Select Health > Alerts > Notification. 2. Select one or more checkboxes of groups in the Notifications group list, and click Delete. The Delete Group dialog box appears.
3. Verify the deletion and click OK.
Resetting the notification group configuration Use the Notification tab to remove all notification groups added and to remove any changes made to the Default group. Procedure 1. Select Health > Alerts > Notification. 2. Select More Tasks > Reset Notification Groups. 3. In the Reset Notification Groups dialog box, click Yes in the verification dialog.
Configuring the daily summary schedule and distribution list Every day, each managed system sends a Daily Alert Summary email to the subscribers configured for the alertssummary.list email group. The Daily Alert Summary email contains current and historical alerts showing messages about non-critical hardware situations and disk space usage numbers that you might want to address soon. A fan failure is an example of a noncritical issue that you might want to address as soon as is reasonably possible. When Support receives the failure notification, they contact you to arrange for component replacement. Procedure 1. Select Health > Alerts > Daily Alert Summary. 2. If the default deliver time of 8 AM is not acceptable, do the following. a. Click Schedule. The Schedule Alert Summary dialog box appears.
b. Use the list boxes to select the hour, minute, and either AM or PM for the summary report.
118 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
c. Click OK. 3. To configure the daily alert subscriber list, do the following. a. Click Configure. The Daily Alert Summary Mailing List dialog box appears.
b. Modify the daily alert subscriber list as follows.
l To add a subscriber, click the + icon, type the email address, and click OK. l To modify an email address, select the checkbox for the subscriber, click the pencil icon, edit the email address, and click OK. l To delete an email address, select the checkbox for the subscriber and click X.
Daily Alert Summary tab The Daily Alert Summary tab allows you to configure an email list of those who want to receive a summary of all system alerts once each day. The people on this list do not receive individual alerts unless they are also added to a notification group.
Table 45 Daily Alert Summary, label descriptions
Delivery Time The delivery time shows the configured time for daily emails. Email List This list displays the email addresses of those who receive the daily emails.
Table 46 Daily Alert Summary tab controls
Configure button Click the Configure button to edit the subscriber email list.
Schedule button Click the Schedule button to configure the time that the daily report is sent.
Enabling and disabling alert notification to EMC You can enable or disable alert notification to EMC without affecting whether or not autosupport reports are sent to EMC. Procedure
1. To view the alert reporting status, select Maintenance > Support > Autosupport. The alert notification status is highlighted in green next to the Real-time alert label in the EMC Support area. Depending on the current configuration, either an Enable or a Disable button appears in the Real-time alert row. 2. To enable alert reporting to EMC, click Enable in the Real-time alert row. 3. To disable alert reporting to EMC, click Disable in the Real-time alert row.
Enabling and disabling alert notification to EMC 119 Managing Data Domain Systems
Testing the alerts email feature Use the Notification tab to send a test email to select notification groups or email addresses. This feature allows you to determine if the system is configured correctly to send alert messages. Procedure 1. To control whether or not a test alert is sent to EMC Data Domain, do the following.
a. Select Maintenance > Support > Autosupport. b. In the Alert Support area, click Enable or Disable to control whether or not the test email is sent . You cannot change the email address.
2. Select Health > Alerts > Notification. 3. Select More Tasks > Send Test Alert. The Send Test Alert dialog box appears.
4. In the Notification Groups list, select groups to receive the test email and click Next. 5. Optionally, add additional email addresses to receive the email. 6. Click Send Now and OK. 7. If you disabled sending of the test alert to EMC Data Domain and you want to enable this feature now, do the following.
a. Select Maintenance > Support > Autosupport. b. In the Alert Support area, click Enable . Results
To test newly added alerts emails for mailer problems, enter: autosupport test email email-addr For example, after adding the email address [email protected] to the list, check the address with the command: autosupport test email [email protected]
EMC Support delivery management Delivery management defines how alerts and autosupport reports are sent to EMC. By default, alerts and autosupport reports are sent to EMC Data Domain Customer Support using the standard (unsecure) email. The ConnectEMC method sends messages in a secure format using email, FTP, or HTTPS. When the ConnectEMC method is used with an EMC Secure Remote Support (ESRS) gateway, one benefit is that one gateway can forward messages from multiple systems, and this allows you to configure network security for only the ESRS gateway instead of for multiple systems.
120 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Selecting standard email delivery to EMC When you select the standard (non-secure) email delivery method, this method applies to both alert and autosupport reporting. Procedure 1. Select Maintenance > Support > Autosupport. 2. Click Configure in the Channel row in the EMC Support area. The Configure EMC Support Delivery dialog appears. The EMC delivery method is displayed after the Channel label in the EMC Support area. 3. In the Channel list box, select Email to datadomain.com. 4. Click OK.
Selecting and configuring ConnectEMC delivery When you select the ConnectEMC delivery method, this method applies to both alert and autosupport reporting. Procedure 1. Select Maintenance > Support > Autosupport. 2. Click Configure in the Channel row in the EMC Support area. The Configure EMC Support Delivery dialog appears. The EMC delivery method is displayed after the Channel label in the EMC Support area. 3. In the Channel list box, select ConnectEMC. The dialog displays the controls for ConnectEMC delivery. 4. To define a delivery method, do the following. a. Click + (plus). b. Select a method in the Method list. The direct method sends messages by secure email directly to EMC. The ESRS- gateway method sends messages to an EMC Secure Remote Support (ESRS) gateway using the specified protocol.
c. Enter a destination IP address or hostname in the Gateway box. d. Click OK. The first delivery method defined becomes the primary delivery method. You can configure only one primary transport. Add additional delivery methods to create alternates to use when the primary delivery method is not available. 5. To change the order of the displayed delivery methods, select a delivery method and use the up and down arrows. The top delivery method in the list is the primary method. 6. In the Email to Default Admin list, select Never, On failure, or Always. This setting determines when the default administrator receives email notification about an alert or autosupport report that is sent to EMC. 7. In the Admin Email box, enter the email address of the administrator who should receive notification when alerts or autosupport reports are sent to EMC. A system administrator email is required to support ConnectEMC delivery.
Selecting standard email delivery to EMC 121 Managing Data Domain Systems
Testing ConnectEMC operation A CLI command allows you to test ConnectEMC operation by sending a dummy file to the system administrator. Before you begin A system administrator address must be configured for the system, and you must disable ConnectEMC reporting before the test. Procedure 1. To disable ConnectEMC reporting, configure the system to use standard email delivery. 2. To test ConnectEMC operation, use the CLI. #support connectemc test Connection Record working successfully After you finish When the test is complete, enable ConnectEMC by configuring the system to use Connect EMC delivery.
Log file management The Data Domain system maintains a set of log files, which can be bundled and sent to EMC Support to assist in troubleshooting any system issues that may arise. Log files cannot be modified or deleted by any user with DD System Manager, but they can be copied from the log directory and managed off of the system.
Log messages on an HA system are preserved on the node where the log file originated.
Log files are rotated weekly. Every Sunday at 0:45 a.m., the system automatically opens new log files for the existing logs and renames the previous files with appended numbers. For example, after the first week of operation, the previous week messages file is renamed messages.1, and new messages are stored in a new messages file. Each numbered file is rolled to the next number each week. For example, after the second week, the file messages.1 is rolled to messages.2. If a messages.2 file already existed, it rolls to messages.3. At the end of the retention period (shown in the table below, the expired log is deleted. For example, an existing messages.9 file is deleted when messages.8 rolls to messages.9. Except as noted in this topic, the log files are stored in /ddvar/log.
Files in the /ddvar directory can be deleted using Linux commands if the Linux user is assigned write permission for that directory.
The set of log files on each system is determined by the features configured on the system and the events that occur. The following table describes the log files that the system can generate.
122 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Table 47 System log files
Log File Description Retention Period
audit.log Messages about user log-in events. 15 weeks
cifs.log Log messages from the CIFS subsystem are logged only in 10 weeks debug/cifs/cifs.log. Size limit of 50 MiB.
messages Messages about general system events, including commands 9 weeks executed.
secure.log Messages regarding user events such as successful and failed 9 weeks logins, user additions and deletions, and password changes. Only Admin role users can view this file.
space.log Messages about disk space usage by system components, and A single file is messages from the clean process. A space use message is kept generated every hour. Each time the clean process runs, it permanently. creates approximately 100 messages. All messages are in There is no log comma-separated-value format with tags you can use to file rotation for separate the disk space messages from the clean process this log. messages. You can use third-party software to analyze either set of messages. The log file uses the following tags.
l CLEAN for data lines from clean operations.
l CLEAN_HEADER for lines that contain headers for the clean operations data lines.
l SPACE for disk space data lines.
l SPACE_HEADER for lines that contain headers for the disk space data lines.
Viewing log files in DD System Manager Use the Logs tab to view and open the system log files in DD System Manager. Procedure
1. Select Maintenance > Logs. The Logs list displays log file names and the size and generation date for each log file.
2. Click a log file name to view its contents. You may be prompted to select an application, such as Notepad.exe, to open the file.
Displaying a log file in the CLI Use the log view command to view a log file in the CLI. Procedure 1. To view a log file in the CLI, use the log view command. With no argument, the command displays the current messages file. 2. When viewing the log, use the up and down arrows to scroll through the file; use the q key to quit; and enter a slash character (/) and a pattern to search through the file.
Viewing log files in DD System Manager 123 Managing Data Domain Systems
The display of the messages file is similar to the following. The last message in the example is an hourly system status message that the Data Domain system generates automatically. The message reports system uptime, the amount of data stored, NFS operations, and the amount of disk space used for data storage (%). The hourly messages go to the system log and to the serial console if one is attached.
# log view Jun 27 12:11:33 localhost rpc.mountd: authenticated unmount request from perfsun-g.emc.com:668 for /ddr/col1/segfs (/ddr/ col1/segfs)
Jun 27 12:28:54 localhost sshd(pam_unix)[998]: session opened for user jsmith10 by (uid=0)
Jun 27 13:00:00 localhost logger: at 1:00pm up 3 days, 3:42, 52324 NFS ops, 84763 GiB data col. (1%)
GiB = Gibibytes = the binary equivalent of Gigabytes.
Learning more about log messages Look up error messages in the Error Message Catalog for your DD OS version. In the log file is text similar to the following.
Jan 31 10:28:11 syrah19 bootbin: NOTICE: MSG-SMTOOL-00006: No replication throttle schedules found: setting throttle to unlimited. The components of the message are as follows.
DateTime Host Process [PID]: Severity: MSG-Module-MessageID: Message Severity levels, in descending order, are: Emergency, Alert, Critical, Error, Warning, Notice, Info, Debug. Procedure 1. Go to the EMC Online Support website at https://support.emc.com, enter Error Message Catalog in the search box, and click the search button. 2. In the results list, locate the catalog for your system and click on the link. 3. User your browser search tool to search for a unique text string in the message. The error message description looks similar to the following display.
ID: MSG-SMTOOL-00006 - Severity: NOTICE - Audience: customer
Message: No replication throttle schedules found: setting throttle to unlimited.
Description: The restorer cannot find a replication throttle schedule. Replication is running with throttle set to unlimited.
Action: To set a replication throttle schedule, run the replication throttle add command.
124 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
4. To resolve an issue, do the recommended action. Based on the example message description, one could run the replication throttle add command to set the throttle.
Saving a copy of log files Save log file copies to another device when you want to archive those files. Use NFS, CIFS mount, or FTP to copy the files to another machine. If using CIFS or NFS, mount /ddvar to your desktop and copy the files from the mount point. The following procedure describes how to use FTP to move files to another machine. Procedure 1. On the Data Domain system, use the adminaccess show ftp command to see whether FTP service is enabled. If the service is disabled, use the command adminaccess enable ftp. 2. On the Data Domain system, use the adminaccess show ftp command to see that the FTP access list includes the IP address of your remote machine. If the address is not in the list, use the command adminaccess add ftp ipaddr. 3. On the remote machine, open a web browser. 4. In the Address box at the top of the web browser, use FTP to access the Data Domain system as shown in the following example.
ftp://Data Domain system_name.yourcompany.com/
Some web browsers do not automatically ask for a login if a machine does not accept anonymous logins. In that case, add a user name and password to the FTP line. For example: ftp://sysadmin:your-pw@Data Domain system_name.yourcompany.com/
5. At the login pop-up, log into the Data Domain system as user sysadmin. 6. On the Data Domain system, you are in the directory just above the log directory. Open the log directory to list the messages files. 7. Copy the file that you want to save. Right-click the file icon and select Copy To Folder from the menu. Choose a location for the file copy. 8. If you want the FTP service disabled on the Data Domain system, after completing the file copy, use SSH to log into the Data Domain system as sysadmin and invoke the command adminaccess disable ftp.
Log message transmission to remote systems Some log messages can be sent from the Data Domain system to other systems. DD OS uses syslog to publish log messages to remote systems. A Data Domain system exports the following facility.priority selectors for log files. For information on managing the selectors and receiving messages on a third-party system, see your vendor-supplied documentation for the receiving system.
l *.notice—Sends all messages at the notice priority and higher. l *.alert—Sends all messages at the alert priority and higher (alerts are included in *.notice).
Saving a copy of log files 125 Managing Data Domain Systems
l kern.*—Sends all kernel messages (kern.info log files). The log host commands manage the process of sending log messages to another system.
Viewing the log file transmission configuration Use the log host show CLI command to view whether log file transmission is enabled and which hosts receive log files. Procedure 1. To display the configuration, enter the log host show command.
# log host show Remote logging is enabled. Remote logging hosts log-server
Enabling and disabling log message transmission You must use CLI commands to enable or disable log message transmission. Procedure 1. To enable sending log messages to other systems, use the log host enable command. 2. To disable sending log messages to other systems, use the log host disable command.
Adding or removing a receiver host You must use CLI commands to add or remove a receiver host. Procedure 1. To add a system to the list that receives Data Domain system log messages, use the log host add command. 2. To remove a system from the list that receives system log messages, use the command: log host del.
The following command adds the system named log-server to the hosts that receive log messages.
log host add log-server
The following command removes the system named log-server from the hosts that receive log messages.
log host del log-server
126 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Remote system power management with IPMI Select DD systems support remote power management using the Intelligent Platform Management Interface (IPMI), and they support remote monitoring of the boot sequence using Serial over LAN (SOL). IPMI power management takes place between an IPMI initiator and an IPMI remote host. The IPMI initiator is the host that controls power on the remote host. To support remote power management from an initiator, the remote host must be configured with an IPMI username and password. The initiator must provide this username and password when attempting to manage power on a remote host. IPMI runs independently of DD OS and allows an IPMI user to manage system power as long as the remote system is connected to a power source and a network. An IP network connection is required between an initiator and a remote system. When properly configured and connected, IPMI management eliminates the need to be physically present to power on or power off a remote system. You can use both DD System Manager and the CLI to configure IPMI users on a remote system. After you configure IPMI on a remote system, you can use IPMI initiator features on another system to log in and manage power.
If a system cannot support IPMI due to hardware or software limitations, DD System Manager displays a notification message when attempting to navigate to a configuration page.
SOL is used to view the boot sequence after a power cycle on a remote system. SOL enables text console data that is normally sent to a serial port or to a directly attached console to be sent over a LAN and displayed by a management host. The DD OS CLI allows you to configure a remote system for SOL and view the remote console output. This feature is supported only in the CLI.
IPMI power removal is provided for emergency situations during which attempts to shut down power using DD OS commands fail. IPMI power removal simply removes power to the system, it does not perform an orderly shutdown of the DD OS file system. The proper way to remove and reapply power is to use the DD OS system reboot command. The proper way to remove system power is to use the DD OS system poweroff command and wait for the command to properly shut down the file system.
IPMI and SOL limitations IPMI and SOL support is limited on some Data Domain systems.
l IPMI is supported on all systems supported by this release except the following systems: DD140, DD610, and DD630. l IPMI user support varies as follows.
n Model DD990: Maximum user IDs = 15. Three default users (NULL, anonymous, root). Maximum user IDs available = 12. n Models DD640, DD4200, DD4500, DD7200, and DD9500: Maximum user IDs = 10. Two default users (NULL, root). Maximum user IDs available = 8.
Remote system power management with IPMI 127 Managing Data Domain Systems
l SOL is supported on the following systems: DD160, DD620, DD640, DD670, DD860, DD890, DD990, DD2200, DD2500 (requires DD OS 5.4.0.6 or later), DD4200, DD4500, DD7200, and DD9500.
User root is not supported for IPMI connections on DD160 systems.
Adding and deleting IPMI users with DD System Manager Each system contains its own list of configured IPMI users, which is used to control access to local power management features. Another system operating as an IPMI initiator can manage remote system power only after providing a valid username and password. To give an IPMI user the authority to manage power on multiple remote systems, you must add that user to each of the remote systems.
The IPMI user list for each remote system is separate from the DD System Manager lists for administrator access and local users. Administrators and local users do not inherit any authorization for IPMI power management.
1. Select Maintenance > IPMI. 2. To add a user, complete the following steps. a. Above the IPMI Users table, click Add. b. In the Add User dialog box, type the user name (16 or less characters) and password in the appropriate boxes (reenter the password in the Verify Password box). c. Click Create. The user entry appears in the IPMI Users table.
3. To delete a user, complete the following steps. a. In the IPMI Users list, select a user and click Delete. b. In the Delete User dialog box, click OK to verify user deletion.
Changing an IPMI user password Change the IPMI user password to prevent use of the old password for power management. Procedure
1. Select Maintenance > IPMI. 2. In the IPMI Users table, select a user, and click Change Password. 3. In the Change Password dialog box, type the password in the appropriate text box and reenter the password in the Verify Password box. 4. Click Update.
128 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Configuring an IPMI port When you configure an IPMI port for a system, you select the port from a network ports list and specify the IP configuration parameters for that port. The selection of IPMI ports displayed is determined by the Data Domain system model. Some systems support one or more dedicated ports, which can be used only for IPMI traffic. Other systems support ports that can be used for both IPMI traffic and all IP traffic supported by the physical interfaces in the Hardware > Ethernet > Interfaces view. Shared ports are not provided on systems that provide dedicated IPMI ports. The port names in the IPMI Network Ports list use the prefix bmc, which represents baseboard management controller. To determine if a port is a dedicated port or shared port, compare the rest of the port name with the ports in the network interface list. If the rest of the IPMI port name matches an interface in the network interface list, the port is a shared port. If the rest of the IPMI port name is different from the names in the network interface list, the port is a dedicated IPMI port.
DD4200, DD4500, and DD7200 systems are an exception to the naming ruled described earlier. On these systems, IPMI port, bmc0a, corresponds to shared port ethMa in the network interface list. EMC recommends that the shared port ethMa be reserved for IPMI traffic and system management traffic (using protocols such as HTTP, Telnet, and SSH). Backup data traffic should be directed to other ports.
When IPMI and nonIPMI IP traffic share an Ethernet port, EMC recommends that you do not use the link aggregation feature on the shared interface because link state changes can interfere with IPMI connectivity. Procedure
1. Select Maintenance > IPMI. The IPMI Configuration area shows the IPMI configuration for the managed system. The Network Ports table lists the ports on which IPMI can be enabled and configured. The IPMI Users table lists the IPMI users who can access the managed system.
Table 48 Network Ports list column descriptions
Port The logical name for a port that supports IPMI communications. Enabled Whether the port is enabled for IPMI (Yes or No). DHCP Whether the port uses DHCP to set its IP address (Yes or No). MAC Address The hardware MAC address for the port. IP Address The port IP address. Netmask The subnet mask for the port. Gateway The gateway IP address for the port.
Configuring an IPMI port 129 Managing Data Domain Systems
Table 49 IPMI Users list column descriptions
User Name The name of a user with authority to power manage the remote system.
2. In the Network Ports table, select a port to configure.
If the IPMI port also supports IP traffic (for administrator access or backup traffic), the interface port must be enabled before you configure IPMI.
3. Above the Network Ports table, click Configure. The Configure Port dialog box appears.
4. Choose how network address information is assigned.
l To collect the IP address, netmask, and gateway configuration from a DHCP server, select Dynamic (DHCP). l To manually define the network configuration, select Static (Manual) and enter the IP address, netmask, and gateway address.
5. Enable a disabled IPMI network port by selecting the network port in the Network Ports table, and clicking Enable. 6. Disable a disabled IPMI network port by selecting the network port in the Network Ports table, and clicking Disable. 7. Click Apply.
Preparing for remote power management and console monitoring with the CLI Remote console monitoring uses the Serial Over Lan (SOL) feature to enable viewing of text-based console output without a serial server. You must use the CLI to set up a system for remote power management and console monitoring. Remote console monitoring is typically used in combination with the ipmi remote power cycle command to view the remote system’s boot sequence. This procedure should be used on every system for which you might want to remotely view the console during the boot sequence. Procedure 1. Connect the console to the system directly or remotely.
l Use the following connectors for a direct connection.
n DIN-type connectors for a PS/2 keyboard n USB-A receptacle port for a USB keyboard n DB15 female connector for a VGA monitor
130 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
Systems DD4200, DD4500, and DD7200 do not support direct connection, including KVM.
l For a serial connection, use a standard DB9 male or micro-DB9 female connector. Systems DD4200, DD4500, and DD7200 provide a female micro-DB9 connector. A null modem cable with male micro-DB9 and standard female DB9 connectors is included for a typical laptop connection. l For a remote IPMI/SOL connection, use the appropriate RJ45 receptacle as follows.
n For DD990 systems, use default port eth0d. n For other systems, use the maintenance or service port. For port locations, refer to the system documentation, such as a hardware overview or installation and setup guide.
2. To support remote console monitoring, use the default BIOS settings. 3. To display the IPMI port name, enter ipmi show config. 4. To enable IPMI, enter ipmi enable . 5. To configure the IPMI port, enter ipmi config port < dhcp | ipaddress ipaddr netmask mask gateway ipaddr >.
If the IPMI port also supports IP traffic (for administrator access or backup traffic), the interface port must be enabled with the net enable command before you configure IPMI.
6. If this is the first time using IPMI, run ipmi user reset to clear IPMI users that may be out of synch between two ports, and to disable default users. 7. To add a new IPMI user, enter ipmi user add user. 8. To set up SOL, do the following: a. Enter system option set console lan.
b. When prompted, enter y to reboot the system.
Managing power with DD System Manager After IPMI is properly set up on a remote system, you can use DD System Manager as an IPMI initiator to log into the remote system, view the power status, and change the power status. Procedure
1. Select Maintenance > IPMI. 2. Click Login to Remote System. The IPMI Power Management dialog box appears.
3. Enter the remote system IPMI IP address or hostname and the IPMI username and password, then click Connect. 4. View the IPMI status.
Managing power with DD System Manager 131 Managing Data Domain Systems
The IPMI Power Management dialog box appears and shows the target system identification and the current power status. The Status area always shows the current status.
The Refresh icon (the blue arrows) next to the status can be used to refresh the configuration status (for example, if the IPMI IP address or user configuration were changed within the last 15 minutes using the CLI commands).
5. To change the IPMI power status, click the appropriate button.
l Power Up—Appears when the remote system is powered off. Click this button to power up the remote system. l Power Down—Appears when the remote system is powered on. Click this button to power down the remote system. l Power Cycle—Appears when the remote system is powered on. Click this button to power cycle the remote system. l Manage Another System—Click this button to log into another remote system for IPMI power management. l Done—Click to close the IPMI Power Management dialog box.
The IPMI Power Down feature does not perform an orderly shutdown of the DD OS. This option can be used if the DD OS hangs and cannot be used to gracefully shutdown a system.
Managing power with the CLI You can manage power on a remote system and start remote console monitoring using the CLI.
The remote system must be properly set up before you can manage power or monitor the system.
Procedure 1. Establish a CLI session on the system from which you want to monitor a remote system. 2. To manage power on the remote system, enter ipmi remote power ipmi-target user user. 3. To begin remote console monitoring, enter ipmi remote console ipmi- target user user.
The user name is an IPMI user name defined for IPMI on the remote system. DD OS user names are not automatically supported by IPMI.
4. To disconnect from a remote console monitoring session and return to the command line, enter the at symbol (@).
132 EMC Data Domain Operating System 5.7 Administration Guide Managing Data Domain Systems
5. To terminate remote console monitoring, enter the tilde symbol (~).
Managing power with the CLI 133 Managing Data Domain Systems
134 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 4 Monitoring Data Domain Systems
This chapter includes:
l Viewing individual system status and identity information. 136 l Health Alerts panel. 138 l Viewing and clearing current alerts. 139 l Viewing the alerts history. 140 l Viewing hardware component status. 141 l Viewing system statistics. 144 l Viewing active users. 145 l History report management. 146 l Viewing the Task Log. 152 l Viewing the system High Availability status. 153
Monitoring Data Domain Systems 135 Monitoring Data Domain Systems
Viewing individual system status and identity information The Dashboard area displays summary information and status for alerts, the file system, licensed services, and hardware enclosures. The Maintenance area displays additional system information, including the system uptime and system and chassis serial numbers. The system name, software version, and user information appear in the footer at all times. Procedure
1. To view system dashboard, select Home > Dashboard. Figure 4 System dashboard
2. To view the system uptime and identity information, select Maintenance > System. The system uptime and identification information appears in the System area.
Dashboard Alerts area The Dashboard Alerts area shows the count, type, and the text of the most recent alerts in the system for each subsystem (hardware, replication, file system, and others). Click anywhere in the alerts area to display more information on the current alerts.
Table 50 Dashboard Alerts column descriptions
Count A count of the current alerts for the subsystem type specified in the adjacent column. The background color indicates the severity of the alert.
Type The subsystem that generated the alert.
136 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Table 50 Dashboard Alerts column descriptions (continued)
Most recent alerts The text of the most recent alert for the subsystem type specified in the adjacent column
Dashboard File System area The Dashboard File System area displays statistics for the entire file system. Click anywhere in the File System area to display more information.
Table 51 File System area label descriptions
Status The current status of the file system.
X.Xx The average compression reduction factor for the file system.
Used The total file system space being used.
Data Written: Pre-compression The data quantity received by the system prior to compression.
Data Written: Post-compression The data quantity stored on the system after compression.
Dashboard Services area The Dashboard Services area displays the status of the replication, VTL, CIFS, NFS, DD Boost, and vDisk services. Click on a service to display detailed information about that service.
Table 52 Services area column descriptions
Left column The left column lists the services that may be used on the system. These service can include replication, VTL, CIFS, NFS, DD Boost, vDisk.
Right column The right column shows the operational status of the service. For most services, the status is enabled, disabled, or not licensed. The replication service row displays the number of replication contexts that are in normal, warning, and error states. A color coded box displays green for normal operation, yellow for warning situations, or red when errors are present).
Dashboard HA Readiness area In high-availability (HA) systems, the HA panel indicates whether the system can fail over from the active node to the standby node if necessary. You can click on the HA panel to navigate to the High Availability section under HEALTH.
Dashboard File System area 137 Monitoring Data Domain Systems
Dashboard Hardware area The Dashboard Hardware area displays the status of the system enclosures and drives. Click anywhere in the Hardware area to display more information on these components.
Table 53 Hardware area label descriptions
Enclosures The enclosure icons display the number of enclosures operating in the normal (green checkmark) and degraded (red X) states.
Storage The storage icons display the number of disk drives operating in the normal (green checkmark), spare (green +), or failed (red X) state.
Maintenance System area The Maintenance System area displays the system model number, DD OS version, system uptime, and system and chassis serial numbers.
Table 54 System area label descriptions
Model Number The model number is the number assigned to the EMC Data Domain system.
Version The version is the DD OS version and build number of the software running on the system.
System Uptime The system uptime displays how long the system has been running since the last system start. The time in parenthesis indicates when the system uptime was last updated.
System Serial The system serial number is the serial number assigned to the system. On newer No. systems, such as DD4500 and DD7200, the system serial number is independent of the chassis serial number and remains the same during many types of maintenance events, including chassis replacements. On legacy systems, such as DD990 and earlier, the system serial number is set to the chassis serial number.
Chassis Serial The chassis serial number is the serial number on the current system chassis. No.
Health Alerts panel Alerts are messages from system services and subsystems that report system events. The Health > Alerts panel displays tabs that allow you to view current and non-current alerts, the configured alert notification groups, and the configuration for those who want to receive daily alert summary reports. Alerts are also sent as SNMP traps. See the MIB Quick Reference Guide or the SNMP MIB for the full list of traps.
138 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Viewing and clearing current alerts The Current Alerts tab displays a list of all the current alerts and can display detailed information for a selected alert. An alert is automatically removed from the Current Alerts list when the underlying situation is corrected or when manually cleared. Procedure
1. To view all of the current alerts, select Health > Alerts > Current Alerts. 2. To limit the number of entries in the current alert list, do the following. a. In the Filter By area, select a Severity and Class to expose only alerts that pertain to those choices. b. Click Update. All alerts not matching the Severity and Class are removed from the list.
3. To display additional information for a specific alert in the Details area, click the alert in the list. 4. To clear an alert, select the alert checkbox in the list and click Clear. A cleared alert no longer appears in the current alerts list, but it can be found in the alerts history list.
5. To remove filtering and return to the full listing of current alerts, click Reset.
Current Alerts tab The Current Alerts tab displays a list of alerts and detailed information about a selected alert.
Table 55 Alerts list, column label descriptions
Message The alert message text. Severity The level of seriousness of the alert. For example, warning, critical, info, or emergency. Date The time and date the alert occurred. Class The subsystem where the alert occurred. Object The physical component where the alert is occurring.
Table 56 Details area, row label descriptions
Name A textual identifier for the alert. Message The alert message text. Severity The level of seriousness of the alert. For example, warning, critical, info, emergency. Class The subsystem and device where the alert occurred.
Viewing and clearing current alerts 139 Monitoring Data Domain Systems
Table 56 Details area, row label descriptions (continued)
Date The time and date the alert occurred. Object ID The physical component where the alert is occurring. Event ID An event identifier. Tenant Units Lists affected tenant units. Description More descriptive information about the alert. Action A suggestion to remedy the alert. Object Info Additional information about the affected object. SNMP OID SNMP object ID.
Viewing the alerts history The Alerts History tab displays a list of all the cleared alerts and can display detailed information for a selected alert. Procedure 1. To view all of the alerts history, select Health > Alerts > Alerts History. 2. To limit the number of entries in the current alert list, do the following. a. In the Filter By area, select a Severity and Class to expose only alerts that pertain to those choices. b. Click Update. All alerts not matching the Severity and Class are removed from the list.
3. To display additional information for a specific alert in the Details area, click the alert in the list. 4. To remove filtering and return to the full listing of cleared alerts, click Reset.
Alerts History tab The Alerts History tab displays a list of cleared alerts and details about a selected alert.
Table 57 Alerts list, column label descriptions
Message The alert message text. Severity The level of seriousness of the alert. For example, warning, critical, info, or emergency. Date The time and date the alert occurred. Class The subsystem where the alert occurred. Object The physical component where the alert is occurring. Status Whether the status is posted or cleared. A posted alert is not cleared.
140 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Table 58 Details area, row label descriptions
Name A textual identifier for the alert. Message The alert message text. Severity The level of seriousness of the alert. For example, warning, critical, info, emergency, Class The subsystem and device where the alert occurred. Date The time and date the alert occurred. Object ID The physical component where the alert is occurring. Event ID An event identifier. Tenant Units Lists affected tenant units.
Additional Information More descriptive information about the alert. Status Whether the status is posted or cleared. A posted alert is not cleared. Description More descriptive information about the alert. Action A suggestion to remedy the alert.
Viewing hardware component status The Hardware Chassis panel displays a block drawing of each enclosure in a system, including the chassis serial number and the enclosure status. Within each block drawing are the enclosure components, such as disks, fans, power supplies, NVRAM, CPUs, and memory. The components that appear depend upon the system model. On systems running DD OS 5.5.1 and later, the system serial number is also displayed. On newer systems, such as DD4500 and DD7200, the system serial number is independent of the chassis serial number and remains the same during many types of maintenance events, including chassis replacements. On legacy systems, such as DD990 and earlier, the system serial number is set to the chassis serial number. Procedure
1. Select Hardware > Chassis. The Chassis view shows the system enclosures. Enclosure 1 is the system controller, and the rest of the enclosures appear below Enclosure 1. Components with problems show yellow (warning) or red (error); otherwise, the component displays OK.
2. Hover the cursor over a component to see detailed status.
Viewing hardware component status 141 Monitoring Data Domain Systems
Fan status Fans are numbered and correspond to their location in the chassis. Hover over a system fan to display a tooltip for that device.
Table 59 Fan tooltip, column label descriptions
Description The name of the fan. Level The current operating speed range (Low, Medium, High). The operating speed changes depending on the temperature inside the chassis. Status The health of the fan.
Temperature status Data Domain systems and some components are configured to operate within a specific temperature range, which is defined by a temperature profile that is not configurable. Hover over the Temperature box to display the temperature tooltip.
Table 60 Temperature tooltip, column label descriptions
Description The location within the chassis being measured. The components listed depend on the model and are often shown as abbreviations. Some examples are:
l CPU 0 Temp (Central Processing Unit)
l MLB Temp 1 (main logic board)
l BP middle temp (backplane)
l LP temp (low profile of I/O riser FRU)
l FHFL temp (full height full length of I/O riser FRU)
l FP temp (front panel)
C/F The C/F column displays temperature in degrees Celsius and Fahrenheit. When the description for a CPU specifies relative (CPU n Relative), this column displays the number of degrees that each CPU is below the maximum allowable temperature and the actual temperature for the interior of the chassis (chassis ambient). Status Shows the temperature status:
l OK—The temperature is acceptable
l Critical—The temperature is higher than the shutdown temperature.
l Warning—The temperature is higher than the warning temperature (but lower than the shutdown temperature).
l Dash (-) —No temperature thresholds are configured for this component, so there is no status to report.
142 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Table 60 Temperature tooltip, column label descriptions (continued)
Power supply status The tooltip shows the status of the power supply (OK or DEGRADED if a power supply is absent or failed). You can also look at the back panel of the enclosure and check the LED for each power supply to identify those that need replacing.
PCI slot status The PCI slots shown in the chassis view indicate the number of PCI slots and the numbers of each slot. Tooltips provide component status for each card in a PCI slot. For example, the tooltip for one NVRAM card model displays the memory size, temperature data, and battery levels.
NVRAM status Hover over NVRAM to display information about the Non-Volatile RAM, batteries, and other components.
Table 61 NVRAM tooltip, dolumn label descriptions
Component The items in the component list depend on the NVRAM installed in the system and can include the following items.
l Firmware version
l Flash controller error counts
l Board temperature
l CPU temperature
l Battery number (The number of batteries depends on the system type.)
l Current slot number for NVRAM
C/F Displays the temperature for select components in the Celsius/ Fahrenheit format. Value Values are provided for select components and describe the following.
l Firmware version number
l Memory size value in the displayed units
l Error counts for memory, PCI, and controller
l Flash controller error counts sorted in the following groups: configuration errors (Cfg Err), panic conditions (Panic), Bus Hang, bad block warnings (Bad Blk Warn), backup errors (Bkup Err), and restore errors (Rstr Err)
Power supply status 143 Monitoring Data Domain Systems
Table 61 NVRAM tooltip, dolumn label descriptions (continued)
l Battery information, such percent charged and status (enabled or disabled)
Viewing system statistics The Realtime Charts panel displays up to seven graphs that show real-time subsystem performance statistics, such as CPU usage and disk traffic. Procedure
1. Select Home > Realtime Charts. The Performance Graphs area displays the currently selected graphs. 2. To change the selection of graphs to display, select and clear the checkboxes for graphs in the list box. 3. To view specific data-point information, hover over a graph point. 4. When a graph contains multiple data, you can use the checkboxes in the upper-right corner of the graph to select what to display. For example, if Read is not selected in the upper right of the disk activity graph, only write data is graphed. Results Each graph shows usage over the last 200 seconds. Click Pause to temporarily stop the display. Click Resume to restart it and show points missed during the pause.
Performance statistics graphs The performance statistics graphs display statistics for key system components and features.
DD Boost Active Connections The DD Boost Active Connections graph displays the number of active DD Boost connections for each of the past 200 seconds. Separate lines within the graph display counts for Read (recovery) connections and Write (backup) connections. DD Boost Data Throughput The DD Boost Data Throughput graph displays the bytes/second transferred for each of the past 200 seconds. Separate lines within the graph display the rates for data read from the system by DD Boost clients and data written to the system by DD Boost clients. Disk The Disk graph displays the amount of data in the appropriate unit of measurement based on the data received, such as KiB or MiB per second, going to and from all disks in the system. File System Operations The File System Operations graph displays the number of operations per second that occurred for each of the past 200 seconds. Separate lines within the graph display the NFS and CIFS operations per second.
144 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Network The Network graph displays the amount of data in the appropriate unit of measurement based on the data received, such as KiB or MiB per second, that passes through each Ethernet connection. One line appears for each Ethernet port. Recent CPU Usage The Recent CPU Usage graph displays the percentage of CPU usage for each of the past 200 seconds. Replication (DD Replicator must be licensed) The Replication graph displays the amount of replication data traveling over the network for each of the last 200 seconds. Separate lines display the In and Out data as follows:
l In: The total number of units of measurement, such as kilobytes per second, received by this side from the other side of the Replicator pair. For the destination, the value includes backup data, replication overhead, and network overhead. For the source, the value includes replication overhead and network overhead.
l Out: The total number of units of measurement, such as kilobytes per second, sent by this side to the other side of the Replicator pair. For the source, the value includes backup data, replication overhead, and network overhead. For the destination, the value includes replication and network overhead.
Viewing active users The Active Users tab displays the names of users who are logged into the system and statistics about the current user sessions. Procedure
1. Select Administration > Access > Active Users. The Active Users list appears and displays information for each user.
Table 62 Active Users list, column label descriptions
Name User name of the logged-in user. Idle Time since last activity of user. Last Login From System from which the user logged in. Last Login Time Datestamp of when user logged in. TTY Terminal notation for login. GUI appears for DD System Manager users.
To manage local users, click Go to Local Users.
Viewing active users 145 Monitoring Data Domain Systems
History report management DD System Manager enables you to generate reports to track space usage on a Data Domain system for up to the previous two years. You can also generate reports to help understand replication progress, and view daily and cumulative reports on the file system. The Reports view is divided into two sections. The upper section lets you create the various types of reports. The lower section lets you view and manage saved reports. Reports display in a table format, and as charts, depending on the type of report. You can select a report for a specific Data Domain system and provide a specific time period. The reports display historical data, not real-time data. After the report is generated, the charts remain static and do not update. Examples of the types of information you can get from the reports include:
l The amount of data that was backed up to the system and the amount of de- duplication that was achieved l Estimates of when the Data Domain system will be full, based on weekly space usage trends l Backup and compression utilization based on selected intervals l Historical cleaning performance, including duration of cleaning cycle, amount of space that can be cleaned, and amount of space that was reclaimed l Amount of WAN bandwidth used by replication, for source and destination, and if bandwidth is sufficient to meet replication requirements l System performance and resource utilization
Types of reports The New Report area lists the types of reports you can generate on your system.
Replication reports can only be created if the system contains a replication license and a valid replication context is configured.
File System Cumulative Space Usage report The File System Cumulative Space Usage Report displays 3 charts that detail space usage on the system during the specified duration. This report is used to analyze how much data is backed up, the amount of deduplication performed, and how much space is consumed.
Table 63 File System—Usage chart label descriptions
Data Written (GiB) The amount of data written before compression. This is indicated by a purple shaded area on the report. Time The timeline for data that was written. The time displayed on this report changes based upon the Duration selection when the chart was created. Total Compression Factor The total compression factor reports the compression ratio.
146 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Table 63 File System—Usage chart label descriptions (continued)
Table 64 File System—Consumption chart label descriptions
Used (GiB) The amount of space used after compression. Time The date the data was written. The time displayed on this report changes based upon the Duration selection when the chart was created. Used (Post Comp) The amount of storage used after compression. Usage Trend The dotted black line shows the storage usage trend. When the line reaches the red line at the top, the storage is almost full. Capacity The total capacity on a Data Domain system.
Cleaning Cleaning is the Cleaning cycle (start and end time for each cleaning cycle). Administrators can use this information to choose the best time for space cleaning the best throttle setting.
Table 65 File System Weekly Cumulative Capacity chart label descriptions
Date (or Time for 24 hour report) The last day of each week, based on the criteria set for the report. In reports, a 24-hour period ranges from noon-to-noon. Data Written (Pre-Comp) The cumulative data written before compression for the specified time period. Used (Post-Comp) The cumulative data written after compression for the specified time period. Compression Factor The total compression factor. This is indicated by a black line on the report.
File System Daily Space Usage report The File System Daily Space Usage report displays five charts that detail space usage during the specified duration. This report is used to analyze daily activities.
Table 66 File System Daily Space Usage chart label descriptions
Space Used (GiB) The amount of space used. Post-comp is red shaded area. Pre- Comp is purple shaded area. Time The date the data was written. Compression Factor The total compression factor. This is indicated by a black square on the report.
Types of reports 147 Monitoring Data Domain Systems
Table 67 File System Daily Capacity Utilization chart label descriptions
Date The date the data was written. Data Written (Pre-Comp) The amount of data written pre-compression. Used (Post-Comp) The amount of storage used after compression. Total Compression Factor The total compression factor.
Table 68 File System Weekly Capacity Utilization chart label descriptions
Start Date The first day of the week for this summary.
End Date The last day of the week for this summary. Available Total amount of storage available. Consumed Total amount of storage used. Data (Post -Comp) The cumulative data written before compression for the specified time period. Replication (Post-Comp) The cumulative data written after compression for the specified time period. Overhead Extra space used for non-data storage. Reclaimed by Cleaning The total space reclaimed after cleaning.
Table 69 File System Compression Summary chart label descriptions
Time The period of data collection for this report. Data Written (Pre-Comp) The amount of data written pre-compression. Used (Post-Comp) The amount of storage used after compression. Total Compression Factor The total compression factor.
Table 70 File System Cleaning Activity chart label descriptions
Start Time The time the cleaning activity started. End Time The time the cleaning activity finished. Duration (Hours) The total time required for cleaning in hours. Space Reclaimed The space reclaimed in Gibibytes (GiB).
148 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Replication Status report The Replication Status report displays three charts that provide the status of the current replication job running on the system. This report is used to provide a snapshot of what is happening for all replication contexts to help understand the overall replication status on a Data Domain System.
Table 71 Replication Context Summary chart label descriptions
ID The Replication Context identification. Source Source system name. Destination Destination system name. Type Type of replication context: MTree, Directory, Collection, or Pool. Status Replication status types include: Error, Normal. Sync as of Time Time and date stamp of last sync. Estimated Completion The estimated time the replication should be complete. Pre-Comp Remaining The amount of pre-compressed data to be replicated. This only applies to Collection type. Post-Comp Remaining The amount of post-compressed data to be replicated. This only applies to Directory and Pool types.
Table 72 Replication Context Error Status chart label descriptions
ID The Replication Context identification. Source Source system name. Destination Destination system name. Type Replication context type: Directory or Pool. Status Replication status types include: Error, Normal, and Warning. Description Description of the error.
Table 73 Replication Destination Space Availability chart label descriptions
Destination Destination system name. Space Availability (GiB) Total amount of storage available.
Types of reports 149 Monitoring Data Domain Systems
Replication Summary report The Replication Summary report provides performance information about a system’s overall network in-and-out usage for replication, as well as per context levels over a specified duration. You select the contexts to be analyzed from a list.
Table 74 Replication Summary report label descriptions
Network In (MiB) The amount of data entering the system. Network In is indicated by a thin green line. Network Out (MiB) The amount of data sent from the system. Network Out is indicated by a thick orange line. Time The date on which the data was written. Pre-Comp Remaining (MiB) The amount of pre-compressed data to be replicated. Pre-Comp Remaining is indicated by a blue line.
Creating a report Create a new report when you want to evaluate file system or replication data collected in the past. Procedure
1. Select Reports > Management. The information panel displays a new report area and a list of saved reports.
2. Click a report type in the New Report area. 3. Select additional options for the report based on the type of report.
l Duration— Last 4 Weeks, Last 7 Days, or Custom l Contexts—Available contexts for working with the Replication Summary report
In reports, the duration of a 24-hour day ranges from noon-to-noon.
4. If you select Custom for the duration, enter Start and End Date in the additional fields. 5. Click Create. The report appears in a separate browser window and is added to the end of the Saved Reports list.
If the report does not display, verify the option to block pop-up windows is enabled on your browser.
150 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Viewing saved reports Saved reports remain available for viewing until deleted. Procedure
1. Select Reports > Management. The Save Reports area lists the available reports.
2. Select the report you want to view. 3. Click View. The report appears in a new browser window.
If the report does not appear, verify the option to block pop-up windows is enabled on your browser.
Printing saved reports After you create a report, you can print it using the controls in your browser window. Procedure
1. Select Reports > Management. 2. Select the report you want to view in the Saved Reports area. 3. Click View. The report appears in a new browser window.
4. In the browser window, select File > Print.
Deleting saved reports You can delete any of the reports that appear in the Saved Reports area. Procedure
1. Select Reports > Maintenance. 2. Select the report you want to delete in the Saved Reports area. You can select multiple reports to delete. Click the checkbox in the heading to select all reports. 3. Click Delete. A warning dialog box asks if you are sure you want to delete the selected reports.
4. Click OK, and click Close.
Renaming saved reports DD System Manager automatically names each report as it is created. You can rename any of the reports that appear in the Saved Reports area. Procedure
1. Select Reports > Management. 2. Select the report you want to rename in the Saved Reports area.
Viewing saved reports 151 Monitoring Data Domain Systems
3. Click Rename. The Rename Report dialog box appears.
4. Type a name for your report in the New Report Name box. It is a good idea to give the report a simple, descriptive name you can easily recognize.
Viewing the Task Log The Task Log displays a list of currently running jobs, such as, replication or system upgrades. DD System Manager can manage multiple systems and can initiate tasks on those systems. If a task is initiated on a remote system, the progress of that task is tracked in the management station task log, not in the remote system task log. Procedure
1. Select Health > Jobs. The Tasks view appears.
2. Select a filter by which to display the Task Log from the Filter By list box. You can select All, In Progress, Failed, or Completed. The Tasks view displays the status of all tasks based on the filter you select and refreshes every 60 seconds.
3. To manually refresh the Tasks list, do either of the following.
l Click Update to update the task log. l Click Reset to display all tasks and remove any filters that were set.
4. To display detailed information about a task, select the task in the task list.
Table 75 Detailed Information, label descriptions
System The system name. Task Description A description of the task. Status The status of the task (completed, failed, or in progress). Start Time The date and time the task started. End Time The date and time the task ended. Error Message An applicable error message, if any.
152 EMC Data Domain Operating System 5.7 Administration Guide Monitoring Data Domain Systems
Viewing the system High Availability status You can use the High Availability panel to see detailed information about the HA status of the system and whether the system can perform failover if necessary. Procedure 1. Select Health > High Availability on the System Manager. The Health High Availability screen appears. A green check mark indicates the system is operating normally and ready for failover. The screen shows the active node, which is typically Node 0. 2. Hover the cursor over a node to see its status. The node is highlighted in blue if it is active. 3. Click the drop-down menu in the banner if you want to change the view from the active node to the standby node, which is typically Node 1.
High Availability status The Health High Availability (HA) view informs you about the system status using a diagram of the nodes and their connected storage. In addition, you can also see any current alerts as well as detailed information about the system. You can determine if the active node and the storage are operational by hovering the cursor over them. Each is highlighted in blue when operating normally. The standby node should appear gray. You can also filter the alerts table by clicking on a component. Only alerts related to the selected components will be displayed.
Viewing the system High Availability status 153 Monitoring Data Domain Systems
Figure 5 Health/High Availability indicators
Table 76 High Availability indicators
HA System bar Displays a green check mark when the system is operating normally and ready for failover.
Failover to Node 0 Allows you to manually fail over to the standby node.
Take Node 1 Offline Allows you to take the active node offline if necessary.
System Information Lists the Data Domain system model, the system type, the version of the Data Domain operating system version in use, and the applied HA license.
HA Manager Displays the nodes, their attached storage, the HA interconnect, and the cabling.
Severity Indicates the severity of any alerts that could impact the system's HA status.
Component Indicates which component is affected.
Class Indicates the class of the alert received such as hardware, environment, and others.
Post Time Indicates the time and date the alert was posted.
154 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 5 File System
This chapter includes:
l File system overview. 156 l Monitoring file system usage. 161 l Managing file system operations. 169 l Fast copy operations. 177
File System 155 File System
File system overview Learn how to use the file system.
How the file system stores data Data Domain storage capacity is best managed by keeping multiple backups and 20% empty space to accommodate backups until the next cleaning. Space use is primarily affected by the size and compressibility of data, and the retention period. A Data Domain system is designed as a very reliable online system for backups and archive data. As new backups are added to the system, old backups are aged out. Such removals are normally done under the control of backup or archive software based on the configured retention period. When backup software expires or deletes an old backup from a Data Domain system, the space on the Data Domain system becomes available only after the Data Domain system cleans the data of the expired backups from disk. A good way to manage space on a Data Domain system is to retain as many online backups as possible with some empty space (about 20% of total space available) to comfortably accommodate backups until the next scheduled cleaning, which runs once a week by default. Some storage capacity is used by Data Domain systems for internal indexes and other metadata. The amount of storage used over time for metadata depends on the type of data stored and the sizes of the stored files. With two otherwise identical systems, one system may, over time, reserve more space for metadata and have less space for actual backup data than the other if different data sets are sent to each system. Space utilization on a Data Domain system is primarily affected by: l The size and compressibility of the backup data. l The retention period specified in the backup software. High levels of compression result when backing up datasets with many duplicates and retaining them for long periods of time.
How the file system reports space usage All DD System Manager windows and system commands display storage capacity using base 2 calculations. For example, a command that displays 1 GiB of disk space as used reports 230 bytes = 1,073,741,824 bytes. l 1 KiB = 210 = 1024 bytes l 1 MiB = 220 = 1,048,576 bytes l 1 GiB = 230 = 1,073,741,824 bytes l 1 TiB = 240 = 1,099,511,627,776 bytes
How the file system uses compression The file system uses compression to optimize available disk space when storing data, so disk space is calculated two ways: physical and logical. (See the section regarding types of compression.) Physical space is the actual disk space used on the Data Domain system. Logical space is the amount of uncompressed data written to the system. The file system space reporting tools (DD System Manager graphs and filesys show space command, or the alias df) show both physical and logical space. These tools also report the size and amounts of used and available space.
156 EMC Data Domain Operating System 5.7 Administration Guide File System
When a Data Domain system is mounted, the usual tools for displaying a file system’s physical use of space can be used. The Data Domain system generates warning messages as the file system approaches its maximum capacity. The following information about data compression gives guidelines for disk use over time. The amount of disk space used over time by a Data Domain system depends on: l The size of the initial full backup. l The number of additional backups (incremental and full) retained over time. l The rate of growth of the backup dataset. l The change rate of data. For data sets with typical rates of change and growth, data compression generally matches the following guidelines: l For the first full backup to a Data Domain system, the compression factor is generally 3:1. l Each incremental backup to the initial full backup has a compression factor generally in the range of 6:1. l The next full backup has a compression factor of about 60:1. Over time, with a schedule of weekly full and daily incremental backups, the aggregate compression factor for all the data is about 20:1. The compression factor is lower for incremental-only data or for backups with less duplicate data. Compression is higher when all backups are full backups.
Types of compression Data Domain compresses data at two levels: global and local. Global compression compares received data to data already stored on disks. Duplicate data does not need to be stored again, while data that is new is locally compressed before being written to disk. Local Compression A Data Domain system uses a local compression algorithm developed specifically to maximize throughput as data is written to disk. The default algorithm (lz) allows shorter backup windows for backup jobs but uses more space. Local compression options provide a trade-off between slower performance and space usage. To change compression, see the section regarding changing local compression. After you change the compression, all new writes use the new compression type. Existing data is converted to the new compression type during cleaning. It takes several rounds of cleaning to recompress all of the data that existed before the compression change. The initial cleaning after the compression change might take longer than usual. Whenever you change the compression type, carefully monitor the system for a week or two to verify that it is working properly.
How the file system implements data integrity Multiple layers of data verification are performed by the DD OS file system on data received from backup applications to ensure that data is written correctly to the Data Domain system disks. This ensures the data can be retrieved without error. The DD OS is purpose-built for data protection and it is architecturally designed for data invulnerability. There are four critical areas of focus, described in the following sections.
End-to-end verification End-to-end checks protect all file system data and metadata. As data comes into the system, a strong checksum is computed. The data is deduplicated and stored in the file
How the file system implements data integrity 157 File System
system. After all data is flushed to disk, it is read back, and re-checksummed. The checksums are compared to verify that both the data and the file system metadata are stored correctly.
Fault avoidance and containment Data Domain uses a log-structured file system that never overwrites or updates existing data. New data is always written in new containers and appended to existing old containers. The old containers and references remain in place and are safe even in the face of software bugs or hardware faults that may occur when storing new backups.
Continuous fault detection and healing Continuous fault detection and healing protects against storage system faults. The system periodically rechecks the integrity of the RAID stripes, and uses the redundancy of the RAID system to heal any faults. During a read, data integrity is re-verified and any errors are healed on the fly.
File system recoverability Data is written in a self-describing format. The file system can be re-created, if necessary, by scanning the log and rebuilding it from the metadata stored with the data.
How the file system reclaims storage space with file system cleaning When your backup application (such as NetBackup or NetWorker) expires data, the data is marked by the Data Domain system for deletion. However, the data is not deleted immediately; it is removed during a cleaning operation.
l During the cleaning operation, the file system is available for all normal operations including backup (write) and restore (read). l Although cleaning uses a significant amount of system resources, cleaning is self- throttling and gives up system resources in the presence of user traffic. l Data Domain recommends running a cleaning operation after the first full backup to a Data Domain system. The initial local compression on a full backup is generally a factor of 1.5 to 2.5. An immediate cleaning operation gives additional compression by another factor of 1.15 to 1.2 and reclaims a corresponding amount of disk space. l When the cleaning operation finishes, a message is sent to the system log giving the percentage of storage space that was reclaimed. A default schedule runs the cleaning operation every Tuesday at 6 a.m. (tue 0600). You can change the schedule or you can run the operation manually (see the section regarding modifying a cleaning schedule). Data Domain recommends running the cleaning operation once a week. Any operation that disables the file system, or shuts down a Data Domain system during a cleaning operation (such as a system power-off or reboot) aborts the cleaning operation. The cleaning operation does not immediately restart when the system restarts. You can manually restart the cleaning operation or wait until the next scheduled cleaning operation. With MTree replication, If a file is created and deleted while a snapshot is being replicated, then the next snapshot will not have any information about this file, and the system will not replicate any content associated with this file. Directory replication will replicate both the create and delete, even though they happen close to each other.
158 EMC Data Domain Operating System 5.7 Administration Guide File System
With the replication log that directory replication uses, operations like deletions, renaming, and so on, execute as a single stream. This can reduce the replication throughput. The use of snapshots by MTree replication avoids this problem.
Supported interfaces Interfaces supported by the file system.
l NFS l CIFS l DD Boost l VTL
Supported backup software Guidance for setting up backup software and backup servers to use with a Data Domain systems is available at support.emc.com.
Data streams sent to a Data Domain system For optimal performance, Data Domain recommends limits on simultaneous streams between Data Domain systems and your backup servers. A data stream, in the context of the following table, refers to a large byte stream associated with sequential file access, such as a write stream to a backup file or a read stream from a restore image. A Replication source or destination stream refers to a directory replication operation or a DD Boost file replication stream associated with a file replication operation.
Table 77 Data streams sent to a Data Domain system
Model RAM/ NVRAM Backup Backup Repla Repla dest Mixed write read source streams streams streams streams
DD140, DD160, 4 GB or 6 GB / 16 4 15 20 w File System. The File System view has a File System panel and six tabs which are described in detail in this section.
About the file system panel Display the file system state and the clean status. State The State area contains an Enable/Disable button and the shows the working state of the file system: l Enabled and running—and the latest consecutive length of time the file system has been enabled and running. l Disabled and shutdown. l Enabling and disabling—in the process of becoming enabled or disabled. l Destroying—if the file system is being deleted. l Error—if there is an error condition, such as a problem initializing the file system. Clean Status The Clean Status area contains a Start Cleaning/Stop Cleaning button and shows the date the last cleaning operation occurred, or the current cleaning status if the cleaning operation is currently running. For example:
Cleaning finished at 2009/01/13 06:00:43 or, if the file system is disabled, shows:
Data Movement Status The Data Movement Status area contains Start/Stop buttons and shows the date the last data movement operation finished, the number of files copied, and the amount of data copied. The system displays a Start button when the data movement operation is available, and a Stop when a data movement operation is running. Space Reclamation Status The Space Reclamation Status area displays the amount of space reclaimed after deleting data in the retention tier. If space reclamation can be run, it shows a Start button. If it is already running, the system displays Stop and Suspend buttons. If the system was running previously and was suspended, it displays Stop and Resume buttons. Click More Information to see space reclamation detailed information: cycle name, start and end time, effective run time, percent completed, units reclaimed, space freed, and total space freed. Storage Migration Status The Storage Migration Status area displays the Storage Migration state including: Starting, Migrating, Paused by User, and so on. The state None indicates that no shelf migration has run on the system or there is no currently running migration. If migration
162 EMC Data Domain Operating System 5.7 Administration Guide File System
has previously run, the system displays the finish time of the last-run migration. If an error occurs during migration, the system displays the error message. If migration is running, the system displays a link to the storage page. Physical Capacity Measurement Status The Physical Capacity Measurement Status area contains an Enable button when physical capacity measurement status is disabled. When enabled, the system displays Disable and View buttons. Click View to see currently running physical capacity measurements: MTree, priority, submit time, start time, and duration.
About the file system summary view Click the Summary tab to view file system statistics.
Space usage Display the amount of disk space available and used by the file system components, based on the last cleaning. l The /data:post-comp line shows amounts for compressed data in the /data directory. l The /ddvar line shows amounts for log and core files. (Remove old logs and core files to free space in this area.)
You can also delete core files from the /ddvar/var/core or the /ddvar/ext directory if it exists.
For both of these, the following amounts are shown in real numbers and in the color- coded graph as described in About the Space Usage View (File System).
l Size—The amount of total physical disk space available for data. l Used—the actual physical space used for compressed data. Warning messages go to the system log and an email alert is generated when the use reaches 90%, 95%, and 100%. At 100%, the Data Domain system accepts no more data from backup servers. If the Used amount is always high, check the cleaning schedule to see how often the cleaning operation runs automatically. Then use the modifying a cleaning schedule procedure to run the operation more often. Also consider reducing the data retention period or splitting off a portion of the backup data to another Data Domain system.
l Available (GiB)—The total amount of space available for data storage. This figure can change because an internal index may expand as the Data Domain system fills with data. The index expansion takes space from the Avail GiB amount. l Cleanable (GiB)—The amount of space that could be reclaimed if a cleaning were run. The second Space Usage panel shows the compression factors:
l Currently Used—The amounts currently in use by the file system. l Written in Last 24 Hours—The compression activity over the last day. For both of these areas, the following is shown:
l Pre-Compression (GiB)—Data written before compression. l Post-Compression (GiB)—Storage used after compression. l Global-Comp Factor—(Pre-Compression / (Size after global compression). l Local-Comp Factor—(Size after global compression) / Post-Compression).
Accessing the file system view 163 File System
l Total-Comp Factor (Reduction %)—Pre-Comp / Post-Comp. l Reduction %—[(Pre-Comp - Post-Comp) / Pre-Comp] * 100.
About the retention units view Display the retention unit and its state, status, and size. The Retention Units view on the File System page is shown only when the optional DD Extended Retention license is enabled. This view lists the retention unit and shows its state (new, sealed, or target), its status (disabled or ready), and its size. If the unit has been sealed, meaning no more data can be added, the date that it was sealed is given. Select the diamond symbol to the right of a column heading to sort the order of the values in reverse.
About the configuration view Display and change system options as well as the current cleaning schedule.
Table 78 Option settings and descriptions
Options settings Description
Local Compression Type The type of local compression in use.
l See the section regarding types of compression for an overview.
l See the section regarding changing local compression
Report Replica as Writable How applications see a replica.
l See the section regarding changing read-only settings
Marker Type Backup software markers (tape markers, tag headers, or other names are used) in data streams. See the section regarding tape marker settings Staging Reserve Manage disk staging.
l See the section regarding working with disk staging
l See the section regarding configuring disk staging
Table 79 Cleaning schedule settings
Cleaning schedule Description settings
Time The date time cleaning operations run.
l See the section regarding modifying a cleaning schedule
Throttle The system resources allocation.
l See the section regarding throttling the cleaning operation
164 EMC Data Domain Operating System 5.7 Administration Guide File System
Table 80 Physical capacity measurement settings
Physical capacity Description measurement settings
Throttle See the section regarding setting the physical capacity measurement throttle. Cache Physical Capacity Cache initialization cleans up the caches and enhances the measuring speed.
Table 81 Data movement policy settings
Data movement policy Description settings
File Age Threshold When data movement starts, all files that have not been modified for the specified threshold number of days will be moved from the active to the retention tier.
Schedule Days and times data is moved. Throttle The percentage of available resources the system uses for data movement. A throttle value of 100% is the default throttle and means that data movement will not be throttled.
About the encryption view Display encryption status, progress, algorithms, and so on.
Table 82 Encryption settings
Encryption Status Status can be one of the following:
l Not licensed—No other information provided.
l Not configured—Encryption is licensed but not configured.
l Enabled—Encryption is enabled and running.
l Disabled—Encryption is disabled.
Encryption Progress View encryption status details for the active tier regarding the application of changes and re-encryption of data. Status can be one of the following:
l Done Click View Details to display the Encryption Status Details dialog that includes the following information for the Active Tier:
Accessing the file system view 165 File System
Table 82 Encryption settings (continued)
l Type (Example: Apply Changes when encryption has already been initiated, or Re-encryption when encryption is a result of compromised data-perhaps a previously destroyed key.)
l Status (Example: Pending)
l Details: (Example: Requested on December xx/xx/xx and will take after the next system clean).
Encryption Algorithm The algorithm used to encrypt the data:
l AES 256-bit (CBC) (default)
l AES 256-bit (GCM) (more secure but slower)
l AES 128-bit (CBC) (not as secure as 256-bit)
l AES 128-bit (GCM) (not as secure as 256-bit) See Changing the Encryption Algorithm for details. Encryption Passphrase When configured, shows as “*****.” To change the passphrase, see Managing the System Passphrase. File System Lock Status The File System Lock status is either:
l Unlocked—The feature is not enabled.
l Locked—The feature is enabled.
Key Management Key Manager Either the internal Data Domain Embedded Key Manager or the optional RSA Data Protection Manager (DPM) Key Manager. Click Configure to switch between key managers (if both are configured), or to modify Key Manager options. Server The name of the RSA Key Manager Server. Server Status Online or offline, or the error messages returned by the RSA Key Manager Server. Key Class A specialized type of security class used by the optional RSA Data Protection Manager (DPM) Key Manager that groups crytopgraphic keys with similar characteristics. The Data Domain system retrieves a key from the RSA server by key class. A key class to be set up to either return the current key, or to generate a new key each time.
The Data Domain system supports only key classes configured to return the current key.
Port The port number of the RSA server. FIPS mode Whether or not the imported host certificate is FIPS compliant. The default mode is enabled. Encryption Keys Lists keys by ID numbers. Shows when a key was created, how long it is valid, its type (RSA DPM Key Manager or the Data Domain internal
166 EMC Data Domain Operating System 5.7 Administration Guide File System
Table 82 Encryption settings (continued)
key), its state (see Working with the RSA DPM Key Manager, DPM Encryption Key States Supported by Data Domain), and the amount of the data encrypted with the key. The system displays the last updated time for key information above the right column. Selected keys in the list can be:
l Synchronized so the list shows new keys added to the RSA server (but are not usable until the file system is restarted).
About the space usage view (file system) Display a visual (but static) representation of data use for the file system at certain points in time. Click the Data Management > File System > Space Usage.
l Click a point on a graph line to display a box with data at that point. l Click Print (at the bottom on the graph) to open the standard Print dialog box. l Click Show in new window to display the graph in a new browser window. The lines of the graph denote measurement for:
l Pre-comp Written—The total amount of data sent to the MTree by backup servers. Pre- compressed data on an MTree is what a backup server sees as the total uncompressed data held by an MTree-as-storage-unit, shown with the Space Used (left) vertical axis of the graph. l Post-comp Used—The total amount of disk storage in use on the MTree, shown with the Space Used (left) vertical axis of the graph. l Comp Factor—The amount of compression the Data Domain system has performed with the data it received (compression ratio), shown with the Compression Factor (right) vertical axis of the graph. Checking Historical Space Usage On the Space Usage graph, clicking an interval (that is, 7d, 30d, 60d, or 120d) on the Duration line above the graph allows you to change the number of days of data shown on the graph, from 7 to 120 days. To see space usage for intervals over 120 days, issue the following command: # filesys show compression [summary | daily | daily-detailed] <[last n
About the consumption view Display space used over time, in relation to total system capacity.
l Click a point on a graph line to display a box with data at that point. l Click Print (at the bottom on the graph) to open the standard Print dialog box. l Click Show in new window to display the graph in a new browser window. The lines of the graph denote measurement for:
Accessing the file system view 167 File System
l Capacity—The total amount of disk storage available for data on the Data Domain system. The amount is shown with the Space Used (left) vertical axis of the graph. Clicking the Capacity checkbox toggles this line on and off. l Post-comp—The total amount of disk storage in use on the Data Domain system. Shown with the Space Used (left) vertical axis of the graph. l Comp Factor—The amount of compression the Data Domain system has performed with the data it received (compression ratio). Shown with the Compression Factor (right) vertical axis of the graph. l Cleaning—A grey diamond is displayed on the chart each time a file system cleaning operation was started. l Data Movement—The amount of disk space moved to the archiving storage area (if the Archive license is enabled). Checking Historical Consumption Usage On the Consumption graph, clicking an interval (that is, 7d, 30d, 60d, 120d) on the Duration line above the graph allows you to change the number of days of data shown on the graph, from 7 to 120 days.
About the daily written view (file system) Display the flow of data over the last 24 hours. The data amounts are shown over time for pre- and post-compression amounts. The system also provides totals for global and local compression amounts, and pre- compression and post-compression amounts.
l Click a point on a graph line to display a box with data at that point. l Click Print (at the bottom on the graph) to open the standard Print dialog box. l Click Show in new window to display the graph in a new browser window. The lines on the graph denote measurements for:
l Pre-Comp Written—The total amount of data written to the file system by backup servers. Pre-compressed data on the file system is what a backup server sees as the total uncompressed data held by the file system. l Post-Comp Written—The total amount of data written to the file system after compression has been performed, as shown in GiBs. l Total Comp Factor—The total amount of compression the Data Domain system has performed with the data it received (compression ratio), shown with the Total Compression Factor (right) vertical axis of the graph. Checking Historical Written Data On the Daily Written graph, clicking an interval (that is, 7d, 30d, 60d, or 120d) on the Duration line above the graph allows you to change the number of days of data shown on the graph, from 7 to 120 days. Below the Daily Written graph, the following totals display for the current duration value:
l Pre-Comp Written l Post-Comp Written l Global-Comp Factor l Local-Comp Factor l Total-Comp Factor (Reduction)
168 EMC Data Domain Operating System 5.7 Administration Guide File System
When the file system is full or nearly full Data Domain systems have three progressive levels of being full. As each level is reached, more operations are progressively disallowed. At each level, deleting data and then performing a file system cleaning operation makes disk space available.
The process of deleting files and removing snapshots does not immediately reclaim disk space, the next cleaning operation reclaims the space.
l Level 1—At the first level of fullness, no more new data can be written to the file system. An informative out of space alert is generated. Remedy—Delete unneeded datasets, reduce the retention period, delete snapshots, and perform a file system cleaning operation. l Level 2—At the second level of fullness, files cannot be deleted. This is because deleting files also require free space but the system has so little free space available that it cannot even delete files. Remedy—Expire snapshots and perform a file system cleaning operation. l Level 3—At the third and final level of fullness, attempts to expire snapshots, delete files, or write new data fail. Remedy—Perform a file system cleaning operation to free enough space to at least delete some files or expire some snapshots and then rerun cleaning.
Monitor the space usage with email alerts Alerts are generated when the file system is at 90%, 95%, and 100% full. To send these alerts, add the user to the alert emailing list.
To join the alert email list, see Viewing and Clearing Alerts.
Managing file system operations This section describes file system cleaning, sanitization, and performing basic operations.
Performing basic operations Basic file system operations include enabling and disabling the file system, and in the rare occasion, destroying a file system.
Creating the file system Create a file system from the Data Management > File System page using the More Tasks menu. There are three reasons to create a file system:
l For a new Data Domain system. l When a system is started after a clean installation. l After a file system has been destroyed. To create the file system:
Managing file system operations 169 File System
Procedure 1. Verify that storage has been installed and configured (see the section on viewing system storage information for more information). If the system does not meet this prerequisite, a warning message is displayed. Install and configure the storage before attempting to create the file system. 2. Select Data Management > File System > Data Management > More Tasks > Create File System. The File System Create dialog box shows the approximate size of the file system. Check Enable file system after creation to start using this file system as soon as it is created. Click Next. A summary displays the file system size and whether the file system is to be automatically enabled. Click Back if you want to change the enable the file system option. Clicking Finish starts the file system creation. A progress bar measures the file system creation’s progress. A check mark indicates that a step of the procedure has completed. When a check mark Completed is displayed, click OK.
Enabling or disabling the file system The option to enable or disable the file system is dependent on the current state of the file system—if its enabled, you can disable it and vice versa.
l Enabling the file system allows Data Domain system operations to begin. This ability is available to administrative users only. l Disabling the file system halts all Data Domain system operations, including cleaning. This ability is available to administrative users only.
Disabling the file system when a backup application is sending data to the system can cause the backup process to fail. Some backup software applications are able to recover by restarting where they left off when they are able to successfully resume copying files; others might fail, leaving the user with an incomplete backup.
1. Select Data Managment > File System. 2. In the File System panel, click Enable or Disable in the State area. 3. Click OK and Close.
Expanding the file system You might need to expand the size of a file system if the suggestions given in "When the File System Is Full or Nearly Full" do not clear enough space for normal operations. A file system may not be expandable, however, for these reasons:
l The file system is not enabled. l There are no unused disks or enclosures in the Active or Retention tiers. l An expanded storage license is not installed. l There are not enough capacity licenses installed. To expand the file system:
170 EMC Data Domain Operating System 5.7 Administration Guide File System
1. Select Data Managment > File System > More Tasks > Expand Capacity. The Expand File System Capacity window displays the current capacity of the file system (active or retention tier) as well as how much additional storage space is available for expansion.
l If enough capacity is available for expansion requirements, continue to 5 on page 171. l If capacity needs to be added, continue with the next step.
If you are configuring a DD2200 system, click the Add License link in the Expand File System Capacity window and enter the CONTROLLER-COD license to expand storage to 7.5 TB, and the EXPANDED-STORAGE license to expand storage to 13.18 TB, and click Expand.
File system capacity can be expanded only if the physical disks are installed on the system and file system is enabled.
2. Click Configure to allocate existing storage to the file system. The Configure Storage dialog box is displayed.
3. In the Available Storage area, click the checkboxes of the storage devices to use and click Add to Tier. System storage must be moved from the Available Storage area to the Active Tier storage area before the file system can be created.
4. Click OK and Close in the progress dialog box. 5. Click Finish to expand the file system into the available storage.
Destroying the file system Destroying the file system should be done only under the direction of Customer Support. This action deletes all data in the Data Domain file system, including virtual tapes. Deleted data is not recoverable. This operation also removes Replication configuration settings. This operation is used when it is necessary to clean out existing data, to create a new collection replication destination, or to replace a collection source, or for security reasons because the system is being removed from operation.
The optional Write zeros to disk operation writes zeros to all file system disks, effectively removing all traces of data. If the Data Domain system contains a large amount of data, this operation can take many hours, or a day, to complete.
As this is a destructive procedure, this operation is available to administrative users only.
Performing basic operations 171 File System
1. Select Data Managment > File System > More Tasks > Distroy. 2. In the Destroy File System dialog box, enter the sysadmin password (it is the only accepted password). 3. Optionally, click the checkbox for Write zeros to disk to completely remove data. 4. Click OK.
Performing cleaning This section describes how to start, stop, and modify cleaning schedules.
Starting cleaning manually To immediately start a cleaning operation. Procedure
1. Select Data Managment > File System > Start Cleaning in the Clean Status area. The system displays the Start File System Clean dialog box.
2. In the Throttle Percentage text box, enter a system throttle amount. This is the percentage of CPU usage dedicated to cleaning. The default is 50 percent. 3. Click OK. The Start File System Clean dialog box allows you to monitor the cleaning operation progress. 4. Click Close to exit the progress dialog box.
Stopping cleaning manually To immediately stop a cleaning operation.
Stopping the cleaning process cancels all cleaning progress for the cleaning run.
1. Select Data Managment > File System > Stop Cleaning. 2. In the File System panel, click Stop Cleaning in the Clean Status area. 3. Click OK.
Modifying a cleaning schedule Use the Modify Schedule dialog of the GUI to modify a cleaning schedule. Procedure 1. Select Data ManagementFile SystemConfiguration. 2. In the Clean Schedule area, click Edit. 3. Select the cleaning frequency, such as daily, bi-weekly, or monthly. Weekly is the default. 4. Enter the start day of the week, time, and throttle percentage.
172 EMC Data Domain Operating System 5.7 Administration Guide File System
Modify the throttle percentage only if necessary. Cleaning leverages all resources when they are idle.
Throttling the cleaning operation Throttle settings to change the amount of system resources used by the cleaning process. Use the Modifying a Cleaning Schedule section to change the system resources used by the cleaning process.
Performing sanitization To comply with government guidelines, system sanitization, also called data shredding, must be performed when classified or sensitive data is written to any system that is not approved to store such data. When an incident occurs, the system administrator must take immediate action to thoroughly eradicate the data that was accidentally written. The goal is to effectively restore the storage device to a state as if the event never occurred. If the data leakage is with sensitive data, the entire storage will need to be sanitized using EMC Professional Services' Secure Data erasure practice. The Data Domain sanitization command exists to enable the administrator to delete files at the logical level, whether a backup set or individual files. Deleting a file in most file systems consists of just flagging the file or deleting references to the data on disk, freeing up the physical space to be consumed at a later time. However, this simple action introduces the problem of leaving behind a residual representation of underlying data physically on disks. Deduplicated storage environments are not immune to this problem. Shredding data in a system implies eliminating the residual representation of that data and thus the possibility that the file may be accessible after it has been shredded. Data Domain's sanitization approach ensures is compliant with the 2007 versions of Department of Defense (DoD) 5220.22 of the following specifications:
l US Department of Defense 5220.22-M Clearing and Sanitization Matrix l National Institute of Systems and Technology (NIST) Special Publication 800-88 Guidelines for Media Sanitization
Sanitizing deduplicated data Data Domain systems sanitize data in place, in its native deduplicated state. Deduplication storage systems extract common data patterns from files sent to the system and store only unique copies of these patterns, referencing all the redundant instances. Because these data patterns or segments may potentially be shared among many files in the system, the sanitization process must first determine whether each of the segments of the contaminated file are shared with a clean file and then erase only those segments that are not shared, along with any contaminated metadata. All storage tiers, caches, unused capacity, and free space are cleared so that every copy of every segment that belongs exclusively to the deleted files is eradicated. The system reclaims and overwrites all of the storage occupied by these segments to effectively restore the storage device to a state as if the contaminated files never existed in that system.
Performing sanitization 173 File System
Sanitization level 1: data clearing or shredding If the data you need to remove is unclassified, as defined in the "US Department of Defense 5220.22-M Clearing and Sanitization Matrix," Level 1 sanitization can be used to overwrite the affected storage once. This provides the basis for handling most data shredding and system sanitization cases. The Data Domain system sanitization feature ensures that every copy of every segment that belongs only to erased files is overwritten using a single-pass zerotization mechanism. Clean data in the system being sanitized is online and available to users. Procedure 1. Delete the contaminated files or backups through the backup software or corresponding client. In the case of backups, be sure to manage the backup software appropriately to ensure that related files on that image are reconciled, catalog records are managed as required, and so forth. 2. Run the system sanitize start command on the contaminated Data Domain system to cause all previously used space in it to be overwritten once (see the figure below). 3. Wait for the affected system to be sanitized. Sanitization can be monitored by using the system sanitize watch command. If the affected Data Domain system has replication enabled, all the systems containing replicas need to be processed in a similar manner. Depending on how much data exists in the system and how it is distributed, the system sanitize command could take some time. However, during this time, all clean data in the system is available to users.
Sanitization level 2: full system sanitization If the data you need to remove is classified, as defined in the "US Department of Defense 5220.22-M Clearing and Sanitization Matrix," Level 2 sanitization, or full system sanitization, is now required. Data Domain recommends Blancco for multi-pass overwrites with any overwrite pattern and a certificate. This provides the basis for handling universal Department of Defense requirements where complete system sanitization is required. For more information, go to: https://www.emc.com/auth/rcoll/servicekitdocument/ cp_datadomaindataerase_psbasddde.pdf
Modifying basic settings Change the type of compression used, marker types, Replica write status, and Staging Reserve percentage, as described in this section.
Changing local compression Use the Local Compression Type area of the GUI to select a new compression type.
Do not change the type of local compression unless it is necessary.
174 EMC Data Domain Operating System 5.7 Administration Guide File System
1. Select Data Management > File System > Configuration. 2. In the Options area, click Edit. The Modify Settings dialog box displays.
3. In the Local Compression Type area, click the drop-down list and select a new compression type.
Table 83 Compression type
none Do not compress data. lz The default algorithm that gives the best throughput. Data Domain recommends the lz option.
gzfast A zip-style compression that uses less space for compressed data, but more CPU cycles (twice as much as lz). Gzfast is the recommended alternative for sites that want more compression at the cost of lower performance. gz A zip-style compression that uses the least amount of space for data storage (10% to 20% less than lz on average; however, some datasets get much higher compression). This also uses the most CPU cycles (up to five times as much as lz). The gz compression type is commonly used for nearline storage applications in which performance requirements are low.
4. Click OK at the confirmation dialog box. 5. Click Close to exit the status dialog box.
Changing read-only settings Change the replica to writable. Some backup applications must see the replica as writable to do a restore or vault operation from the replica. Procedure
1. Select Data Management > File System > Configuration. 2. In the Options area, click Edit. 3. In the Report Replica as Writable area, click the Enable checkbox. 4. Click OK. 5. Click Close to exit the status dialog box.
Working with disk staging Disk staging enables a Data Domain system to serve as a staging device, where the system is viewed as a basic disk via a CIFS share or NFS mount point. Disk staging can be used in conjunction with your backup software, such as Symantec’s NetBackup (NBU) and EMC’s NetWorker, it does not require a license, and is disabled by default.
Modifying basic settings 175 File System
The VTL feature is not required or supported when the Data Domain system is used as a Disk Staging device.
The reason that some backup applications use disk staging devices is to enable tape drives to stream continuously. After the data is copied to tape, it is retained on disk for as long as space is available. Should a restore be needed from a recent backup, more than likely the data is still on disk and can be restored from it more conveniently than from tape. When the disk fills up, old backups can be deleted to make space. This delete-on- demand policy maximizes the use of the disk. In normal operation, the Data Domain System does not reclaim space from deleted files until a cleaning operation is done. This is not compatible with backup software that operates in a staging mode, which expects space to be reclaimed when files are deleted. When you configure disk staging, you reserve a percentage of the total space—typically 20 to 30 percent—in order to allow the system to simulate the immediate freeing of space. The amount of available space is reduced by the amount of the staging reserve. When the amount of data stored uses all of the available space, the system is full. However, whenever a file is deleted, the system estimates the amount of space that will be recovered by cleaning and borrows from the staging reserve to increase the available space by that amount. When a cleaning operation runs, the space is actually recovered and the reserve restored to its initial size. Since the amount of space made available by deleting files is only an estimate, the actual space reclaimed by cleaning may not match the estimate. The goal of disk staging is to configure enough reserve so that you do not run out before cleaning is scheduled to run.
Configuring disk staging Enable disk staging and specify the staging reserve percentage. Procedure
1. Select Data Management > File System > Configuration. 2. In the Options area, click Edit. 3. In the Staging Reserve panel, select the Enable checkbox. 4. Enter a value in the % of Total Space text box. This value represents the percentage of the total disk space to be reserved for disk staging, typically 20 to 30 percent.
Tape marker settings Backup software from some vendors insert markers (tape markers, tag headers, or other names are used) in all data streams (both file system and VTL backups) sent to a Data Domain system. Markers can significantly degrade data compression on a Data Domain system. As such, the default marker type auto is set and cannot be changed by the user. If this setting is not compatible with your backup software, contact your contracted support provider.
176 EMC Data Domain Operating System 5.7 Administration Guide File System
For information about how applications work in a Data Domain environment, see How EMC Data Domain Systems Integrate into the Storage Environment. You can use these matrices and integration guides to troubleshoot vendor-related issues.
Fast copy operations A fast copy operation clones files and directory trees of a source directory to a target directory on a Data Domain system. The force option allows the destination directory to be overwritten if it exists. Executing the fast copy operation displays a progress status dialog box.
A fast copy operation makes the destination equal to the source, but not at a specific time. There are no guarantees that the two are or were ever equal if you change either folder during this operation.
Performing a fast copy operation Copy a file or directory tree from a Data Domain system source directory to another destination on the Data Domain system. Procedure
1. Select Data Management > File System > More Tasks > Fast Copy. 2. In the Source text box, enter the pathname of the directory where the data to be copied resides. For example, /data/col1/backup/.snapshot/snapshot-name/ dir1.
col1 uses a lower case L followed by the number 1.
3. In the Destination text box, enter the pathname of the directory where the data will be copied to. For example, /data/col1/backup/dir2. This destination directory must be empty, or the operation fails.
l If the Destination directory exists, click the checkbox Overwrite existing destination if it exists.
4. Click OK. 5. In the progress dialog box that appears, click Close to exit.
Fast copy operations 177 File System
178 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 6 MTrees
This chapter includes:
l MTrees overview. 180 l Monitoring MTree usage. 187 l Managing MTree operations. 190
MTrees 179 MTrees
MTrees overview An MTree is a logical partition of the Data Domain file system. You can use MTrees in the following ways: for DD Boost storage units, VTL pools, or an NFS/CIFS share. MTrees allow granular management of snapshots, quotas, and Retention Lock. For systems that have DD Extended Retention and granular management of data migration policies from Active Tier to Retention Tier, MTree operations can be performed on a specific MTree as opposed to the entire file system.
There can be up to the maximum configurable MTrees designated for MTree replication contexts.
MTree limits MTree limits for Data Domain systems
Table 84 Supported MTrees
Data Domain System DD OS Supported Supported concurrently Version configurable MTrees active MTrees
DD9500 5.7 and later 256 256
DD990, DD4200, 5.7 and later 128 128 DD4500, DD7200
All other DD systems 5.7 and later 100 Up to 32 based on the model
DD9500 5.6 100 64
DD990, DD890 5.3 and later 100 Up to 32 based on the model
DD7200, DD4500, 5.4 and later 100 Up to 32 based on the model DD4200
All other DD systems 5.2 and later 100 Up to 14 based on the model
Quotas MTree quotas apply only to the logical data written to the MTree. An administrator can set the storage space restriction for an MTree, Storage Unit, or VTL pool to prevent it from consuming excess space. There are two kinds of quota limits: hard limits and soft limits. You can set either a soft or hard limit or both a soft and hard limit. Both values must be integers, and the soft value must be less than the hard value. When a soft limit is set, an alert is sent when the MTree size exceeds the limit, but data can still be written to it. When a hard limit is set, data cannot be written to the MTree when the hard limit is reached. Therefore, all write operations fail until data is deleted from the MTree. See the section regarding MTree quota configuration for more information.
180 EMC Data Domain Operating System 5.7 Administration Guide MTrees
Quota enforcement Enable or disable quota enforcement.
About the MTree panel Lists all the active MTrees on the system and shows real-time data storage statistics. Information in the overview area is helpful in visualizing space usage trends. Select Data Management > MTree.
l Select a checkbox of an MTree in the list to display details and perform configuration in the Summary view. l Enter text (wildcards are supported) in the Filter By MTree Name field and click Update to list specific MTree names in the list. l Delete filter text and click Reset to return to the default list.
Table 85 MTree overview information
MTree Name The pathname of the MTree. Quota Hard Limit Percentage of hard limit quota used. Quota Soft Limit Percentage of hard limit quota used. Last 24 Hr Pre-Comp (pre- Amount of raw data from the backup application that has been compression) written in the last 24 hours. Last 24 Hr Post-Comp (post- Amount of storage used after compression in the last 24 hours. compression) Last 24 hr Comp Ratio The compression ratio for the last 24 hours. Weekly Avg Post-Comp Average amount of compressed storage used in the last five weeks. Last Week Post-Comp Average amount of compressed storage used in the last seven days. Weekly Avg Comp Ratio The average compression ratio for the last five weeks. Last Week Comp Ratio The average compression ratio for the last seven days.
About the summary view View important file system statistics.
View detail information Select an MTree to view information.
Table 86 MTree detail information for a selected MTree
Full Path The pathname of the MTree.
About the MTree panel 181 MTrees
Table 86 MTree detail information for a selected MTree (continued)
Pre-Comp Used The current amount of raw data from the backup application that has been written to the MTree. Status The status of the MTree (combinations are supported). Status can be:
l RD: Replication destination
l RLCE: Retention Lock Compliance enabled
l RLCD: Retention Lock Compliance disabled
l RLGE: Retention Lock Governance enabled
l RLGD: Retention Lock Governance disabled
Tenant Unit Tenant-unit information displays if the selected MTree belongs to a tenant-unit. References to tenant-units display only when Secure Multi-tenancy is enabled. Quota Quota Enforcement Enabled or Disabled. Pre-Comp Soft Limit Current value. Click Configure to revise the quota limits. Pre-Comp Hard Limit Current value. Click Configure to revise the quota limits. Quota Summary Percentage of Hard Limit used. Protocols CIFS Shared The CIFS share status. Status can be:
l Yes—The MTree or its parent directory is shared.
l Partial—The subdirectory under this MTree is shared.
l No—This MTree and its parent or subdirectory are not shared. Click the CIFS link to go to the CIFS view. NFS Exported The NFS export status. Status can be:
l Yes—The MTree or its parent directory is exported.
l Partial—The subdirectory under this MTree is exported.
l No—This MTree and its parent or subdirectory are not exported. Click the NFS link to go to the NFS view. DD Boost Storage Unit The DD Boost export status. Status can be:
l Yes—The MTree is exported.
l No—This MTree is not exported.
l Unknown—There is no information. Click the DD Boost link to go to the DD Boost view.
182 EMC Data Domain Operating System 5.7 Administration Guide MTrees
Table 86 MTree detail information for a selected MTree (continued)
VTL Pool If applicable, the name of the VTL pool that was converted to an MTree. vDisk Pool vDisk report status. Status can be:
l Unknown— vDisk service is not enabled.
l No— vDisk service is enabled but the MTree is not a vDisk pool.
l Yes— vDisk service is enabled and the MTree is a vDisk pool.
Physical Capacity Measurements Used (Post-Comp) MTree space that is used after compressed data has been ingested.
Compression Global Comp-factor. Last Measurement Time Last time the system measured the MTree. Schedules Number of schedules assigned.
Click Assign to view and assign schedules to the MTree.
l Name: The schedule name.
l Status: Enabled or Disabled
n Normal— Submits a measurement task to the processing queue.
n Urgent— Submits a measurement task to the front of the processing queue.
l Schedule: Time the task runs.
l MTree Assignments: Number of MTrees the schedule is assigned to.
Submitted Measurements Displays the post compression status for the MTree.
Click Measure Now to submit a manual post compression job for the MTree and select a priority for the job.
l 0— No measurement job submitted.
l 1— 1 measurement job running.
l 2— 2 measurement jobs running.
View MTree replication information Display MTree replication configuration. If the selected MTree is configured for replication, summary information about the configuration displays in this area. Otherwise, this area displays No Record Found.
l Click the Replication link to go to the Replication page for configuration and to see additional details.
About the summary view 183 MTrees
Table 87 MTree replication information
Source The source MTree pathname. Destination The destination MTree pathname. Status The status of the MTree replication pair. Status can be Normal, Error, or Warning. Sync As Of The last day and time the replication pair was synchronized.
View MTree snapshot information If the selected MTree is configured for snapshots, summary information about the snapshot configuration displays.
l Click the Snapshots link to go to the Snapshots page to perform configuration or to see additional details. l Click Assign Snapshot Schedules to assign a snapshot schedule to the selected MTree. Select the schedule’s checkbox, and then click OK and Close. To create a snapshot schedule, click Create Snapshot Schedule (see the section about creating a snapshot schedule for instructions).
Table 88 MTree snapshot information
Total Snapshots The total number of snapshots created for this MTree. A total of 750 snapshots can be created for each MTree. Expired The number of snapshots in this MTree that have been marked for deletion, but have not been removed with the clean operation as yet. Unexpired The number of snapshots in this MTree that are marked for keeping. Oldest Snapshot The date of the oldest snapshot for this MTree. Newest Snapshot The date of the newest snapshot for this MTree. Next Scheduled The date of the next scheduled snapshot. Assigned Snapshot The name of the snapshot schedule assigned to this MTree. Schedules
View MTree retention lock information If the selected MTree is configured for one of the Retention Lock software options, summary information about the Retention Lock configuration displays.
For information on how to manage Retention Lock for an MTree, see the section about working with DD retention lock.
184 EMC Data Domain Operating System 5.7 Administration Guide MTrees
Table 89 Retention lock information
Status Indicates whether Retention Lock is enabled or disabled. Retention Period Indicates the minimum and maximum Retention Lock time periods. UUID Shows either:
l the unique identification number generated for an MTree when the MTree is enabled for Retention Lock
l that the Retention Lock on a file in the MTree has been reverted
Enabling and managing DD retention lock settings Use the Retention Lock area of the GUI to modify retention lock periods. Procedure
1. Select Data Management > MTree > Summary. 2. In the Retention Lock area, click Edit. 3. In the Modify Retention Lock dialog box, select Enable to enable Retention Lock on the Data Domain system. 4. Modify the minimum or maximum retention period (the feature must be enabled first), in the Retention Period panel. 5. Select an interval (minutes, hours, days, years). Click Default to show the default values. 6. Click OK. Results After you close the Modify Retention Lock dialog box, updated MTree information is displayed in the Retention Lock summary area.
About the space usage view (MTrees) Display a visual representation of data usage for an MTree at certain points in time. Select Data Management > MTree > Space Usage.
l Click a point on a graph line to display a box with data at that point. l Click Print (at the bottom on the graph) to open the standard Print dialog box. l Click Show in new window to display the graph in a new browser window. The lines of the graph denote measurement for:
l Pre-comp Written—The total amount of data sent to the MTree by backup servers. Pre- compressed data on an MTree is what a backup server sees as the total uncompressed data held by an MTree-as-storage-unit, shown with the Space Used (left) vertical axis of the graph.
About the space usage view (MTrees) 185 MTrees
For the MTrees Space Usage view, the system displays only pre-compressed information. Data can be shared between MTrees so compressed usage for a single MTree cannot be provided.
Checking Historical Space Usage On the Space Usage graph, clicking an interval (that is, 7d, 30d, 60d, or 120d) on the Duration line above the graph allows you to change the number of days of data shown on the graph, from 7 to 120 days. To see space usage for intervals over 120 days, issue the following command: # filesys show compression [summary | daily | daily-detailed] <[last n
About the daily written view (MTrees) Display the flow of data over the last 24 hours. Data amounts are shown over time for pre- and post-compression. It also provides totals for global and local compression amounts, and pre-compression and post-compression amounts.
l Click a point on a graph line to display a box with data at that point. l Click Print (at the bottom on the graph) to open the standard Print dialog box. l Click Show in new window to display the graph in a new browser window. The lines on the graph denote measurements for:
l Pre-Comp Written—The total amount of data written to the MTree by backup servers. Pre-compressed data on an MTree is what a backup server sees as the total uncompressed data held by an MTree -as-storage-unit. l Post-Comp Written—The total amount of data written to the MTree after compression has been performed, as shown in GiBs. l Total Comp Factor—The total amount of compression the Data Domain system has performed with the data it received (compression ratio), shown with the Total Compression Factor (right) vertical axis of the graph. Checking Historical Written Data On the Daily Written graph, clicking an interval (that is, 7d, 30d, 60d, or 120d) on the Duration line above the graph allows you to change the number of days of data shown on the graph, from 7 to 120 days. Below the Daily Written graph, the following totals display for the current duration value:
l Pre-Comp Written l Post-Comp Written l Global-Comp Factor l Local-Comp Factor l Total-Comp Factor
186 EMC Data Domain Operating System 5.7 Administration Guide MTrees
Monitoring MTree usage Display space usage and data written trends for an MTree. Procedure
l Select Data Management > MTtee. The MTree view shows a list of configured MTrees, and when selected in the list, details of the MTree are shown in the Summary tab. The Space Usage and Daily Written tabs show graphs that visually display space usage amounts and data written trends for a selected MTree. The view also contains options that allow MTree configuration for CIFS, NFS, and DD Boost, as well as sections for managing snapshots and Retention Lock for an MTree. The MTree view has an MTree overview panel and three tabs which are described in detail in these sections.
n About the MTree panel on page 181 n About the summary view on page 181 n About the space usage view (MTrees) on page 185 n About the daily written view (MTrees) on page 186
Physical capacity measurement (PCM) provides space usage information for MTrees. For more information about PCM, see the section regarding understanding physical capacity measurement.
Understanding physical capacity measurement Physical capacity measurement (PCM) provides space usage information for a sub-set of storage space. From the System Manager, PCM provides space usage information for MTrees, but from the command line interface you can view space usage information for MTrees, tenants, tenant units, and pathsets. For more information about how to use PCM from the command line, see the EMC Data Domain Operating System Command Reference Guide.
Enabling, disabling, and viewing physical capacity measurement Physical capacity measurement provides space usage information for an MTree. Procedure
1. Select Data Management > File System > File System. The system displays the Summary tab in the File System panel. 2. Click Enable to the right of Physical Capacity Measurement Status to enable PCM. 3. Click View to the right of Physical Capacity Measurement Status to view currently running PCM jobs.
l MTree: The MTree that PCM is measuring. l Priority: The priority (normal or urgent) for the task. l Submit Time: The time the task was requested. l Duration: The length of time PCM ran to accomplish of the task.
Monitoring MTree usage 187 MTrees
4. Click Disable to the right of Physical Capacity Measurement Status to disable PCM and cancel all currently running PCM jobs.
Initializing physical capacity measurement Physical capacity measurement (PCM) initialization is a one-time action that can take place only if PCM is enabled and the cache has not been initialized. It cleans the caches and enhances measuring speed. During the initialization process, you can still manage and run PCM jobs. Procedure
1. Select Data Management > File System > Configuration. 2. Click Initialize under Physical Capacity Measurement to the right of Cache. 3. Click Yes.
Managing physical capacity measurement schedules Create, edit, delete, and view physical capacity measurement schedules. This dialog only displays schedules created for MTrees and schedules that currently have no assignments. Procedure
1. Select Data Management > MTree > Manage Schedules.
l Click Add (+) to create a schedule. l Select a schedule and click Modify (pencil) to edit the schedule. l Select a schedule and click Delete (X) to delete a schedule. 2. Optionally, click the heading names to sort by schedule: Name, Status (Enabled or Disabled) Priority (Urgent or Normal), Schedule (schedule timing), and MTree Assignments (the number of MTrees the schedule is assigned to).
Creating physical capacity measurement schedules Create physical capacity measurement schedules and assign them to MTrees. Procedure
1. Select Data Management > MTree > Manage Schedules. 2. Click Add (+) to create a schedule. 3. Enter the name of the schedule. 4. Select the status:
l Normal: Submits a measurement task to the processing queue. l Urgent: Submits a measurement task to the front of the processing queue.
5. Select how often the schedule triggers a measurement occurrence: every Day, Week, or Month.
l For Day, select the time. l For Week, select the time and day of the week. l For Month, select the time, and days during the month.
6. Select MTree assignments for the schedule (the MTrees that the schedule will apply to): 7. Click Create.
188 EMC Data Domain Operating System 5.7 Administration Guide MTrees
8. Optionally, click on the heading names to sort by schedule: Name, Status (Enabled or Disabled) Priority (Urgent or Normal), Schedule (schedule timing), and MTree Assignments (the number of MTrees the schedule is assigned to).
Editing physical capacity measurement schedules Edit a physical capacity measurement schedule. Procedure
1. Select Data Management > MTree > Manage Schedules. 2. Select a schedule and click Modify (pencil). 3. Modify the schedule and click Save. Schedule options are described in the Creating physical capacity measurement schedules topic. 4. Optionally, click the heading names to sort by schedule: Name, Status (Enabled or Disabled) Priority (Urgent or Normal), Schedule (schedule timing), and MTree Assignments (the number of MTrees the schedule is assigned to).
Assigning physical capacity measurement schedules to an MTree Attach schedules to an MTree. Before you begin Physical capacity measurement (PCM) schedules must be created.
Administrators can assign up to three PCM schedules to an MTree.
1. Select Data Management > MTree > Summary. 2. Select MTrees to assign schedules to. 3. Scroll down to the Physical Capacity Measurements area and click Assign to the right of Schedules. 4. Select schedules to assign to the MTree and click Assign.
Starting physical capacity measurement immediately Start the measurement process as soon as possible. Procedure
1. Select Data Management > MTree > Summary. 2. Scroll down to the Physical Capacity Measurements area and click Measure Now to the right of Submitted Measurements. 3. Select Normal (Submits a measurement task to the processing queue), or Urgent (Submits a measurement task to the front of the processing queue). 4. Click Submit.
Understanding physical capacity measurement 189 MTrees
Setting the physical capacity measurement throttle Set the percentage of system resources that are dedicated to physical capacity measurement. Procedure
1. Select Data Management > File System > Configuration. 2. In the Physical Capacity Measurement area, click Edit to the left of Throttle. 3. Option Description Click Default Enters the 20% system default. Type throttle percent The percentage of system resources that are dedicated to physical capacity measurement.
Managing MTree operations This section describes MTree creation, configuration, how to enable and disable MTree quotas, and so on.
Creating an MTree An MTree is a logical partition of the Data Domain file system. Use MTrees in for DD Boost storage units, VTL pools, or an NFS/CIFS share. MTrees are created in the area /data/col1/mtree_name. Procedure
1. Select Data Management > MTree. 2. In the MTree overview area, click Create. 3. Enter the name of the MTree in the MTree Name text box. MTree names can be up to 50 characters. The following characters are acceptable:
l Upper- and lower-case alphabetical characters: A-Z, a-z l Numbers: 0-9 l Embedded space l comma (,) l period (.), as long as it does not precede the name. l explanation mark (!) l number sign (#) l dollar sign ($) l per cent sign (%) l plus sign (+) l at sign (@) l equal sign (=) l ampersand (&)
190 EMC Data Domain Operating System 5.7 Administration Guide MTrees
l semi-colon (;) l parenthesis [(and)] l square brackets ([and]) l curly brackets () l caret (^) l tilde (~) l apostrophe (unslanted single quotation mark) l single slanted quotation mark (‘)
4. Set storage space restrictions for the MTree to prevent it from consuming excessive space. Enter a soft or hard limit quota setting, or both. With a soft limit, an alert is sent when the MTree size exceeds the limit, but data can still be written to the MTree. Data cannot be written to the MTree when the hard limit is reached.
The quota limits are pre-compressed values. To set quota limits for the MTree, select Set to Specific value and enter the value. Select the unit of measurement: MiB, GiB, TiB, or PiB.
When setting both soft and hard limits, a quota’s soft limit cannot exceed the quota’s hard limit.
5. Click OK. The new MTree displays in the MTree table.
You may need to expand the width of the MTree Name column to see the entire pathname.
Configure and enable/disable MTree quotas Set the storage space restriction for an MTree, Storage Unit, or VTL pool. The Data Management > Quota page shows the administrator how many MTrees have no soft or hard quotas set. For MTrees with quotas set, the page shows the percentage of pre-compressed soft and hard limits used. Consider the following information when managing quotas.
l MTree quotas apply to ingest operations. These quotas can be applied to data on systems that have the DD Extended Retention software, regardless of which tier it resides on, as well as VTL, DD Boost, CIFS, and NFS. l Snapshots are not counted. l Quotas cannot be set on the /data/col1/backup directory.
l The maximum quota value allowed is 4096 PiB.
Configure and enable/disable MTree quotas 191 MTrees
Configure MTree quotas Use the MTree tab or the Quota tab to configure MTree quotas. Procedure 1. Select one of the following menu paths:
l Select Data Management > MTree. l Select Data Management > Quota. 2. Select only one MTree in the MTree tab, or one or more MTrees in the Quota tab.
Quotas cannot be set on the /data/col1/backup directory.
3. In the MTree tab, click the Summary tab, and then click the Configure button in the Quota area. 4. In the Quota tab, click the Configure Quota button. Configuring MTree quotas Enter values for hard and soft quotas and select the unit of measurement. Procedure 1. In the Configure Quota for MTrees dialog box, enter values for hard and soft quotas and select the unit of measurement: MiB, GiB, TiB, or PiB. 2. Click OK.
Deleting an MTree Removes the MTree from the MTree table. The MTree data is deleted at the next cleaning.
Because the MTree and its associated data are not removed until file cleaning is run, you cannot create a new MTree with the same name as a deleted MTree until the deleted MTree is completely removed from the file system by the cleaning operation.
1. Select Data Management > MTree. 2. Select an MTree. 3. In the MTree overview area, click Delete. 4. Click OK at the Warning dialog box. 5. Click Close in the Delete MTree Status dialog box after viewing the progress.
Undeleting an MTree Undelete retrieves a deleted MTree and its data and places it back in the MTree table. An undelete of an MTree retrieves a deleted MTree and its data and places it back in the MTree table. An undelete is possible only if file cleaning has not been run after the MTree was marked for deletion.
192 EMC Data Domain Operating System 5.7 Administration Guide MTrees
You can also use this procedure to undelete a storage unit.
1. Select Data Management > MTree > More Tasks > Undelete. 2. Select the checkboxes of the MTrees you wish to bring back and click OK. 3. Click Close in the Undelete MTree Status dialog box after viewing the progress. The recovered MTree displays in the MTree table.
Renaming an MTree Use the Data Management MTree GUI to rename MTrees. Procedure
1. Select Data Management > MTree. 2. Select an MTree in the MTree table. 3. Select the Summary tab. 4. In the Detailed Information overview area, click Rename. 5. Enter the name of the MTree in the New MTree Name text box. See the section about creating an MTree for a list of allowed characters.
6. Click OK. The renamed MTree displays in the MTree table.
Replicating a system with quotas to one without Replicate a Data Domain system with a DD OS that supports quotas, to a system with a DD OS that does not have quotas.
l A reverse resync, which takes the data from the system without quotas and puts it back in an MTree on the system that has quotas enabled (and which continues to have quotas enabled). l A reverse initialization from the system without quotas, which takes its data and creates a new MTree on the system that supports quotas, but does not have quotas enabled because it was created from data on a system without quotas.
Quotas were introduced as of DD OS 5.2.
Renaming an MTree 193 MTrees
194 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 7 Snapshots
This chapter includes:
l Snapshots overview. 196 l Monitoring snapshots and their schedules. 197 l Managing snapshots. 198 l Managing snapshot schedules. 200 l Recover data from a snapshot. 202
Snapshots 195 Snapshots
Snapshots overview This chapter describes how to use the snapshot feature with MTrees. A snapshot saves a read-only copy (called a snapshot) of a designated MTree at a specific time. You can use a snapshot as a restore point, and you can manage MTree snapshots and schedules and display information about the status of existing snapshots.
Snapshots created on the source Data Domain system are replicated to the destination with collection and MTree replication. It is not possible to create snapshots on a Data Domain system that is a replica for collection replication. It is also not possible to create a snapshot on the destination MTree of MTree replication. Directory replication does not replicate the snapshots, and it requires you to create snapshots separately on the destination system.
Snapshots for the MTree named backup are created in the system directory /data/ col1/backup/.snapshot. Each directory under /data/col1/backup also has a .snapshot directory with the name of each snapshot that includes the directory. Each MTree has the same type of structure, so an MTree named SantaClara would have a system directory /data/col1/SantaClara/.snapshot, and each subdirectory in / data/col1/SantaClara would have a .snapshot directory as well.
The .snapshot directory is not visible if only /data is mounted. When the MTree itself is mounted, the .snapshot directory is visible.
An expired snapshot remains available until the next file system cleaning operation. The maximum number of snapshots allowed per MTree is 750. Warnings are sent when the number of snapshots per MTree reaches 90% of the maximum allowed number (from 675 to 749 snapshots), and an alert is generated when the maximum number is reached. To clear the warning, expire snapshots and then run the file system cleaning operation.
To identify an MTree that is nearing the maximum number of snapshots, check the Snapshots panel of the MTree page regarding viewing MTree snapshot information.
Snapshot retention for an MTree does not take any extra space, but if a snapshot exists and the original file is no longer there, the space cannot be reclaimed.
Snapshots and CIFS Protocol: As of DD OS 5.0, the .snapshot directory is no longer visible in the directory listing in Windows Explorer or DOS CMD shell. You can access the .snapshot directory by entering its name in the Windows Explorer address bar or the DOS CMD shell. For example, \\dd\backup\.snapshot or Z:\.snapshot when Z: is mapped as \\dd\backup).
196 EMC Data Domain Operating System 5.7 Administration Guide Snapshots
Monitoring snapshots and their schedules This section provides detailed and summary information about the status of snapshots and snapshot schedules.
About the snapshots view The topics in this section describe the Snapshot view.
Snapshots overview panel View the total number of snapshots, the number of expired snapshots, unexpired snapshots, and the time of the next cleaning. Select Data Management > Snapshots.
Table 90 Snapshot overview panel information
Total Snapshots (Across all The total number of snapshots, active and expired, on all MTrees in MTrees) the system. Expired The number of snapshots that have been marked for deletion, but have not been removed with the cleaning operation as yet. Unexpired The number of snapshots that are marked for keeping. Next file system clean The date the next scheduled file system cleaning operation will be scheduled performed.
Snapshots view View snapshot information by name, by MTree, creation time, whether it is active, and when it expires. The Snapshots tab displays a list of snapshots and lists the following information.
Table 91 Snapshot information
Selected Mtree A drop-down list that selects the MTree the snapshot operates on. Filter By Items to search for in the list of snapshots that display. Options are:
l Name—Name of the snapshot (wildcards are accepted). l Year—Drop-down list to select the year.
Name The name of the snapshot image. Creation Time The date the snapshot was created. Expires On The date the snapshot expires. Status The status of the snapshot, which can be Expired or blank if the snapshot is active.
Monitoring snapshots and their schedules 197 Snapshots
Schedules view View the days snapshots will be taken, the times, the time they will be retained, and the naming convention.
Table 92 Snapshot schedule information
Name The name of the snapshot schedule. Days The days the snapshots will be taken. Times The time of day the snapshots will be taken. Retention Period The amount of time the snapshot will be retained. Snapshot Name Pattern A string of characters and variables that translate into a snapshot name (for example, scheduled-%Y-%m-%d-%H-%M, which translates to “scheduled-2010-04-12-17-33”).
1. Select a schedule in the Schedules tab. The Detailed Information area appears listing the MTrees that share the same schedule with the selected MTree. 2. Click the Add/Remove button to add or remove MTrees from schedule list.
Managing snapshots This section describes how to manage snapshots.
Creating a snapshot Create a snapshot when an unscheduled snapshot is required. Procedure
1. Select Data Management > Snapshots to open the Snapshots view. 2. In the Snapshots view, click Create. 3. In the Name text field, enter the name of the snapshot. 4. In the MTree(s) area, select a checkbox of one or more MTrees in the Available MTrees panel and click Add. 5. In the Expiration area, select one of these expiration options: a. Never Expire. b. Enter a number for the In text field, and select Days, Weeks, Month, or Years from the drop-down list. The snapshot will be retained until the same time of day as when it is created. c. Enter a date (using the formatmm/dd/yyyy) in the On text field, or click Calendar and click a date. The snapshot will be retained until midnight (00:00, the first minute of the day) of the given date. 6. Click OK and Close.
198 EMC Data Domain Operating System 5.7 Administration Guide Snapshots
Modifying a snapshot expiration date Modify snapshot expiration dates to remove them or extent their life for auditing or compliance. Procedure 1. Select Data ManagementSnapshots to open the Snapshots view. 2. Click the checkbox of the snapshot entry in the list and click Modify Expiration Date.
More than one snapshot can be selected by clicking additional checkboxes.
3. In the Expiration area, select one of the following for the expiration date: a. Never Expire. b. In the In text field, enter a number and select Days, Weeks, Month, or Years from the drop-down list. The snapshot will be retained until the same time of day as when it is created. c. In the On text field, enter a date (using the format mm/dd/yyyy) or click Calendar and click a date. The snapshot will be retained until midnight (00:00, the first minute of the day) of the given date. 4. Click OK.
Renaming a snapshot Use the Snapshot tab to rename a snapshot. Procedure
1. Select Data Management > Snapshots to open the Snapshots view. 2. Select the checkbox of the snapshot entry in the list and click Rename. 3. In the Name text field, enter a new name. 4. Click OK.
Expiring a snapshot Snapshots cannot be deleted. To release disk space, expire snapshots and they will be deleted in the next cleaning cycle after the expiry date. Procedure
1. Select Data Management > Snapshots to open the Snapshots view. 2. Click the checkbox next to snapshot entry in the list and click Expire.
More than one snapshot can be selected by selecting additional checkboxes. The snapshot is marked as Expired in the Status column and will be deleted at the next cleaning operation.
Modifying a snapshot expiration date 199 Snapshots
Managing snapshot schedules Set up and manage a series of snapshots that will be automatically taken at regular intervals (a snapshot schedule). Multiple snapshot schedules can be active at the same time.
If multiple snapshots are scheduled to occur at the same time, only one is retained. Which one is retained is indeterminate, thus only one snapshot should be scheduled for a given time.
Creating a snapshot schedule Create a weekly or monthly snapshot schedule using the Data Management GUI. Procedure
1. Select Data Managment > Snapshots > Schedules to open the Schedules view. 2. Click Create. 3. In the Name text field, enter the name of the schedule. 4. In the Snapshot Name Pattern text box, enter a name pattern. Enter a string of characters and variables that translates to a snapshot name (for example, scheduled-%Y-%m-%d-%H-%m, translates to "scheduled-2012-04-12-17-33"). Use alphabetic characters, numbers, _, -, and variables that translate into current values.
5. Click Validate Pattern & Update Sample. 6. Click Next. 7. Select the date when the schedule will be executed: a. Weekly—Click checkboxes next to the days of the week or select Every Day. b. Monthly—Click the Selected Days option and click the dates on the calendar, or select the Last Day of the Month option. c. Click Next. 8. Select the time of day when the schedule will be executed: a. At Specific Times—Click Add and in the Time dialog that appears, enter the time in the format hh:mm, and click OK. b. In Intervals—Click the drop-down arrows to select the start and end time hh:mm and AM or PM. Click the Interval drop-down arrows to select a number and then the hours or minutes of the interval. c. Click Next. 9. In the Retention Period text entry field, enter a number and click the drop-down arrow to select days, months, or years, and click Next. Schedules must explicitly specify a retention time.
10. Review the parameters in the schedule summary and click Finish to complete the schedule or Back to change any entries.
200 EMC Data Domain Operating System 5.7 Administration Guide Snapshots
11. If an MTree is not associated with the schedule, a warning dialog box asks if you would like to add an MTree to the schedule. Click OK to continue (or Cancel to exit). 12. To assign an MTree to the schedule, in the MTree area, click the checkbox of one or more MTrees in the Available MTrees panel, then click Add and OK.
Naming conventions for snapshots created by a schedule The naming convention for scheduled snapshots is the word scheduled followed by the date when the snapshot is to occur, in the format scheduled-yyyy-mm-dd-hh-mm. For example, scheduled-2009-04-27-13-30. The name “mon_thurs” is the name of a snapshot schedule. Snapshots generated by that schedule might have the names scheduled-2008-03-24-20-00, scheduled-2008-03-25-20-00, etc.
Modifying a snapshot schedule Change the snapshot schedule name, date, and retention period. Procedure 1. In the schedule list, select the schedule and click Modify. 2. In the Name text field, enter the name of the schedule and click Next. Use alphanumeric characters, and the _ and -.
3. Select the date when the schedule is to be executed: a. Weekly—Click checkboxes next to the days of the week or select Every Day. b. Monthly—Click the Selected Days option and click the dates on the calendar, or select the Last Day of the Month option. c. Click Next. 4. Select the time of day when the schedule is to be executed: a. At Specific Times—Click the checkbox of the scheduled time in the Times list and click Edit. In the Times dialog that appears, enter a new time in the format hh:mm, and click OK. Or click Delete to remove the scheduled time. b. In Intervals—Click the drop-down arrows to select the start and end time hh:mm and AM or PM. Click the Interval drop-down arrows to select a number and then the hours or minutes of the interval. c. Click Next. 5. In the Retention Period text entry field, enter a number and click the drop-down arrow to select days, months, or years, and click Next. 6. Review the parameters in the schedule summary and click Finish to complete the schedule or Back to change any entries.
Deleting a snapshot schedule Delete a snapshot schedule from the schedule list. Procedure 1. In the schedule list, click the checkbox to select the schedule and click Delete. 2. In the verification dialog box, click OK and then Close.
Modifying a snapshot schedule 201 Snapshots
Recover data from a snapshot Use the fastcopy operation to retrieve data stored in a snapshot. See the section regarding fast copy operations.
202 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 8 CIFS
This chapter includes:
l CIFS overview. 204 l Configuring SMB signing. 204 l Performing CIFS setup. 205 l Working with shares. 207 l Managing access control. 212 l Monitoring CIFS operation. 217 l Performing CIFS troubleshooting. 220
CIFS overview Common Internet File System (CIFS) clients can have access to the system directories on the Data Domain system. l The /data/col1/backup directory is the destination directory for compressed backup server data. l The /ddvar/core directory contains Data Domain System core and log files (remove old logs and core files to free space in this area).
You can also delete core files from the /dvar or the /ddvar/ext directory if it exists.
Clients, such as backup servers that perform backup and restore operations with a Data Domain System, at the least, need access to the /data/col1/backup directory. Clients that have administrative access need to be able to access the /ddvar/core directory to retrieve core and log files. As part of the initial Data Domain system configuration, CIFS clients were configured to access these directories. This chapter describes how to modify these settings and how to manage data access using the Data DD Manager and the cifs command.
l The DD System Manager Protocols > CIFS page allows you to perform major CIFS operations such as enabling and disabling CIFS, setting authentication, managing shares, and viewing configuration and share information. l The cifs command contains all the options to manage CIFS backup and restores between Windows clients and Data Domain systems, and to display CIFS statistics and status. For complete information about the cifs command, see the EMC Data Domain Operating System Command Reference Guide. l For information about the initial system configuration, see the EMC Data Domain Operating System Initial Configuration Guide. l For information about setting up clients to use the Data Domain system as a server, see the related tuning guide, such as the CIFS Tuning Guide, which is available from the support.emc.com web site. Search for the complete name of the document using the Search EMC Support field.
Configuring SMB signing On a DD OS version that supports it, you can configure the SMB signing feature using the CIFS option called server signing. This feature is disabled by default because it degrades performance. When enabled, SMB signing can cause a 29 percent (reads) to 50 percent (writes) throughput performance drop, although individual system performance will vary. There are three possible values for SMB signing: disabled, auto and mandatory:
l When SMB signing is set to disabled, SMB signing is disabled, this is the default.
204 EMC Data Domain Operating System 5.7 Administration Guide CIFS
l When SMB signing is set to required, SMB signing is required, and both computers in the SMB connection must have SMB signing enabled. SMB Signing CLI Commands cifs option set "server-signing" required Sets server signing to required. cifs option reset "server-signing" Resets server signing to the default (disabled). As a best practice, whenever you change the SMB signing options, disable and then enable (restart) CIFS service using the following CLI commands: cifs disable cifs enable The System Manager interface displays whether the SMB signing option is disabled or set to auto or mandatory. To view this setting in the interface, navigate to: Protocols > CIFS > Configuration tab. In the Options area, the value for the SMB signing option will be disabled, auto or mandatory reflecting the value set using the CLI commands.
Performing CIFS setup This section contains instructions about enabling CIFS services, naming the CIFS server, and so on.
HA systems and CIFS HA systems are compatible with CIFS; however, if a CIFS job is in progress during a failover, the job will need to be restarted. "/ddvar is an ext3 file system, and cannot be shared like a normal MTree-based share. The information in /ddvar will become stale when the active node fails over to the standby node because the filehandles are different on the two nodes. If /ddvar is mounted to access log files or upgrade the system, unmount and remount /ddvar if a failover has occurred since the last time /ddvar was mounted."
Preparing clients for access to Data Domain systems Find EMC documentation online. Procedure 1. Log into the EMC Online Support (support.emc.com) web site. 2. In the Search EMC Support field, enter the name of the document that you are looking for. 3. Select the appropriate document, such as the CIFS and Data Domain Systems Tech Note. 4. Follow the instructions in the document.
Enabling CIFS services Enable the client to access the system using the CIFS protocol. After configuring a client for access to Data Domain systems, enable CIFS services, which allows the client to access the system using the CIFS protocol.
Performing CIFS setup 205 CIFS
Procedure 1. For the Data Domain system that is selected in the DD System Manager Navigation tree, click Protocols > CIFS. 2. In the CIFS Status area, click Enable.
Naming the CIFS server The hostname for the Data Domain system that serves as the CIFS server is set during the system’s initial configuration. To change a CIFS server name, see the procedures in the section regarding setting authentication parameters. A Data Domain system’s hostname should match the name assigned to its IP address, or addresses, in the DNS table. Otherwise authentication, as well as attempts to join a domain, can fail. If you need to change the Data Domain system’s hostname, use the net set hostname command, and also modify the system’s entry in the DNS table. When the Data Domain system acts as a CIFS server, it takes the hostname of the system. For compatibility purposes, it also creates a NetBIOS name. The NetBIOS name is the first component of the hostname in all uppercase letters. For example, the hostname jp9.oasis.local is truncated to the NetBIOS name JP9. The CIFS server responds to both names. You can have the CIFS server respond to different names at the NetBIOS levels by changing the NetBIOS hostname.
Changing the NetBIOS hostname Change the NetBIOS hostname with the CLI. Procedure 1. Display the current NetBIOS name by entering:
# cifs show config 2. Use the cifs set nb-hostname nb-hostname command.
Setting authentication parameters Set the Data Domain authentication parameters for working with CIFS. Click the Configure link in to the left of the Authentication label in the Configuration tab. The system will navigate to the Administration > Access > Authentication tab where you can configure authentication for Active Directory, Kerberos, Workgroups, and NIS.
Setting CIFS options View CIFS configuration, restrict anonymous connections. Procedure
1. Select Protocols > CIFS > Configuration. 2. In the Options area, click Configure Options. 3. To restrict anonymous connections, click the checkbox of the Enable option in the Restrict Anonymous Connections area. 4. In the Log Level area, click the drop-down list to select the level number.
206 EMC Data Domain Operating System 5.7 Administration Guide CIFS
The level is an integer from 1 (one) to 5 (five). One is the default system level that sends the least-detailed level of CIFS-related log messages, five results in the most detail. Log messages are stored in the file /ddvar/log/debug/cifs/cifs.log.
A log level of 5 degrades system performance. Click the Default in the Log Level area after debugging an issue. This sets the level back to 1.
5. In the Server Signing area, select:
l Enabled to enable server signing l Disabled to disable server signing l Required when server signing is required
Disabling CIFS services Prevent clients from accessing the Data Domain system. Procedure
1. Select Protocols > CIFS. 2. In the Status area, click Disable. 3. Click OK. Even after disabling CIFS access, CIFS authentication services continue to run on the Data Domain system. This continuation is required to authenticate active directory domain users for management access.
Working with shares To share data, create shares on the Data Domain system. Shares are administered on the Data Domain system and the CIFS systems.
Creating shares on the Data Domain system When creating shares, you have to assign client access to each directory separately and remove access from each directory separately. For example, a client can be removed from /ddvar and still have access to /data/col1/backup A Data Domain system supports a maximum number of 3000 CIFS shares.1 And 600 simultaneous connections are allowed. However, the maximum number of connections supported is based on system memory. See the section regarding setting the maximum open files on a connection for more information.
If Replication is to be implemented, a Data Domain system can receive backups from both CIFS clients and NFS clients as long as separate directories are used for each. Do not mix CIFS and NFS data in the same directory.
1. May be affected by hardware limitations.
Disabling CIFS services 207 CIFS
1. Select Protocols > CIFS tabs to navigate to the CIFS view. 2. Ensure authentication has been configured, as described in the section regarding setting authentication parameters. 3. On the CIFS client, set shared directory permissions or security options. 4. On the CIFS view, click the Shares tab. 5. Click Create. 6. In the Create Shares dialog box, enter the following information:
Table 93 Shares dialog box information
Share Name A descriptive name for the share.
Directory Path The path to the target directory (for example, /data/col1/ backup/dir1).
col1 uses the lower case letter L followed by the number 1.
Comment A descriptive comment about the share.
The share name can be a maximum of 80 characters and cannot contain the following characters: \ / : * ? " | + [ ] ; , = or extended ASCII characters.
7. Add a client by clicking Add (+) in the Clients area. The Client dialog box appears. Enter the name of the client in the Client text box and click OK. Consider the following when entering the client name.
l No blanks or tabs (white space) characters are allowed. l It is not recommended to use both an asterisk (*) and individual client name or IP address for a given share. When an asterisk (*) is present, any other client entries for that share are not used. l It is not required to use both client name and client IP address for the same client on a given share. Use client names when the client names are defined in the DNS table. l To make share available to all clients, specify an asterisk (*) as the client. All users in the client list can access the share, unless one or more user names are specified, in which case only the listed names can access the share. Repeat this step for each client that you need to configure.
8. In the Max Connections area, select the text box and enter the maximum number of connections to the share that are allowed at one time. The default value of zero (also settable via the Unlimited button) enforces no limit on the number of connections. 9. Click OK.
208 EMC Data Domain Operating System 5.7 Administration Guide CIFS
The newly created share appears at the end of the list of shares, located in the center of the Shares panel.
Modifying a share on a Data Domain system Change share information and connections. Procedure
1. Select Protocols > CIFS > Shares to navigate to the CIFS view, Shares tab. 2. Click the checkbox next the share that you wish to modify in the Share Name list. 3. Click Modify. 4. Modify share information: a. To change the comment, enter new text in the Comment text field. b. To modify a User or Group names, in the User/Group list, click the checkbox of the user or group and click Edit (pencil icon) or Delete (X). To add a user or group, click (+), and in the User/Group dialog box select the Type for User or Group, and enter the user or group name. c. To modify a client name, in the Client list click the checkbox of the client and click Edit (pencil icon) or Delete (X). To add a client, click the Add (+) and add the name in the Client dialog box.
To make the share available to all clients, specify an asterisk (*) as the client. All users in the client list can access the share, unless one or more user names are specified, in which case only the listed names can access the share.
d. Click OK. 5. In the Max Connections area, in the text box, change the maximum number of connections to the share that are allowed at one time. Or select Unlimited to enforce no limit on the number of connections. 6. Click OK.
Creating a share from an existing share Create a share from an existing share and modify the new share if necessary.
User permissions from the existing share are carried over to the new share.
Procedure 1. In the CIFS Shares tab, click the checkbox for the share you wish to use as the source. 2. Click Create From. 3. Modify the share information, as described in the section about modifying a share on a Data Domain system.
Modifying a share on a Data Domain system 209 CIFS
Disabling a share on a Data Domain system Disable one or more existing shares. Procedure 1. In the Shares tab, click the checkbox of the share you wish to disable in the Share Name list. 2. Click Disable. 3. Click Close.
Enabling a share on a Data Domain system Enable one or more existing shares. Procedure 1. In the Shares tab, click the checkbox of the shares you wish to enable in the Share Name list. 2. Click Enable. 3. Click Close.
Deleting a share on a Data Domain system Delete one or more existing shares. Procedure 1. In the Shares tab, click the checkbox of the shares you wish to delete in the Share Name list. 2. Click Delete. The Warning dialog box appears.
3. Click OK. The shares are removed.
Performing MMC administration Use the Microsoft Management Console (MMC) for administration. DD OS supports these MMC features:
l Share management, except for browsing when adding a share, or the changing of the offline settings default, which is a manual procedure. l Session management. l Open file management, except for deleting files.
Connecting to a Data Domain system from a CIFS client Use CIFS to connect to a Data Domain system and create a read-only backup subfolder. Procedure 1. On the Data Domain system CIFS page, verify that CIFS Status shows that CIFS is enabled and running.
210 EMC Data Domain Operating System 5.7 Administration Guide CIFS
2. In the Control Panel, open Administrative Tools and select Computer Management. 3. In the Computer Management dialog box, right-click Computer Management (Local) and select Connect to another computer from the menu. 4. In the Select Computer dialog box, select Another computer and enter the name or IP address for the Data Domain system. 5. Create a \backup subfolder as read-only. For more information, see the section on creating a /data/col1/backup subfolder as read-only. Figure 6 Computer Management dialog box
Creating a \data\col1\backup subfolder as read-only Enter a path, share name, and select permissions. Procedure 1. In the Control Panel, open Administrative Tools and select Computer Management. 2. Right-click Shares in the Shared Folders directory. 3. Select New File Share from the menu. The Create a Shared Folder wizard opens. The computer name should be the name or IP address of the Data Domain system.
4. Enter the path for the Folder to share, for example, enter C:\data\col1\backup \newshare.
5. Enter the Share name, for example, enter newshare. Click Next. 6. For the Share Folder Permissions, selected Administrators have full access. Other users have read-only access. Click Next.
Connecting to a Data Domain system from a CIFS client 211 CIFS
Figure 7 Completing the Create a Shared Folder Wizard
7. The Completing dialog shows that you have successfully shared the folder with all Microsoft Windows clients in the network. Click Finish. The newly created shared folder is listed in the Computer Management dialog box.
Displaying CIFS information Display information about shared folders, sessions, and open files. Procedure 1. In the Control Panel, open Administrative Tools and select Computer Management. 2. Select one of the Shared Folders (Shares, Sessions, or Open Files) in the System Tools directory. Information about shared folders, sessions, and open files is shown in the right panel.
Managing access control Access shared from a Windows client, provide administrative access, and allow access from trusted domain users.
Accessing shares from a Windows client Use the command line to map a share. Procedure
l From the Windows client use this DOS command: net use drive: backup-location
212 EMC Data Domain Operating System 5.7 Administration Guide CIFS
For example, enter:
# \\dd02\backup /USER:dd02\backup22 This command maps the backup share from Data Domain system dd02 to drive H on the Windows system and gives the user named backup22 access to the \\DD_sys\backup directory.
Providing domain users administrative access Use the command line to add CIFS and include the domain name in the ssh instruction. Procedure
l Enter: adminaccess authentication add cifs The SSH, Telnet, or FTP command that accesses the Data Domain system must include, in double quotation marks, the domain name, a backslash, and the user name. For example: C:> ssh "domain2\djones" @dd22
Allowing administrative access to a Data Domain system for domain users Use the command line to map a DD system default group number, and then enable CIFS administrative access. Procedure 1. To map a Data Domain System default group number to a Windows group name that differs from the default group name, use the cifs option set "dd admin group2" ["windows grp-name"] command. The Windows group name is a group (based on one of the user roles—admin, user, or back-up operator) that exists on a Windows domain controller, and you can have up to 50 groups (dd admin group1 to dd admin group50).
For a description of DD OS user roles and Windows groups, see the section about managing Data Domain systems.
2. Enable CIFS administrative access by entering:
adminaccess authentication add cifs
l The default Data Domain System group dd admin group1 is mapped to the Windows group Domain Admins. l You can map the default Data Domain System group dd admin group2 to a Windows group named Data Domain that you create on a Windows domain controller. l Access is available through SSH, Telnet, FTP, HTTP, and HTTPS. l After setting up administrative access to the Data Domain system from the Windows group Data Domain, you must enable CIFS administrative access using the adminaccess command.
Providing domain users administrative access 213 CIFS
Restricting administrative access from Windows Use the command line to prohibit access to users without a DD account. Procedure
l Enter: adminaccess authentication del cifs This command prohibits Windows users access to the Data Domain system if they do not have an account on the Data Domain system.
File access This sections contains information about ACLs, setting DACL and SACL permissions using Windows Explorer, and so on.
NT access control lists Access control lists (ACLs) are enabled by default on the Data Domain system.
Data Domain recommends that you do not disable NTFS ACLs once they have been enabled. Contact Data Domain Support prior to disabling NTFS ACLs.
Default ACL Permissions The default permissions, which are assigned to new objects created through the CIFS protocol when ACLs are enabled, depend on the status of the parent directory. There are three different possibilities:
l The parent directory has no ACL because it was created through NFS protocol. l The parent directory has an inheritable ACL, either because it was created through the CIFS protocol or because ACL had been explicitly set. The inherited ACL is set on new objects. l The parent directory has an ACL, but it is not inheritable. The permissions are as follows:
Table 94 Permissions
Type Name Permission Apply To
Allow SYSTEM Full control This folder only Allow CREATOR OWNER Full control This folder only
CREATOR OWNER is replaced by the user creating the file/folder for normal users and by Administrators for administrative users.
Permissions for a New Object when the Parent Directory Has No ACL The permissions are as follows:
l BUILTIN\Administrators:(OI)(CI)F l NT AUTHORITY\SYSTEM:(OI)(CI)F l CREATOR OWNER:(OI)(CI)(IO)F
214 EMC Data Domain Operating System 5.7 Administration Guide CIFS
l BUILTIN\Users:(OI)(CI)R l BUILTIN\Users:(CI)(special access:)FILE_APPEND_DATA l BUILTIN\Users:(CI)(IO)(special access:)FILE_WRITE_DATA l Everyone:(OI)(CI)R These permissions are described in more detail as follows:
Table 95 Permissions Detail
Type Name Permission Apply To
Allow Administrators Full control This folder, subfolders, and files Allow SYSTEM Full control This folder, subfolders, and files Allow CREATOR OWNER Full control Subfolders and files only
Allow Users Read & execute This folder, subfolders, and files Allow Users Create subfolders This folder and subfolders only Allow Users Create files Subfolders only Allow Everyone Read & execute This folder, subfolders, and files
Setting ACL Permissions and Security Windows-based backup and restore tools such as NetBackup can be used to back up DACL- and SACL-protected files to the Data Domain system, and to restore them from the Data Domain system. Granular and Complex Permissions (DACL) You can set granular and complex permissions (DACL) on any file or folder object within the DDFS file systems, either through using Windows commands such as cacls, xcacls, xcopy and scopy, or through the CIFS protocol using the Windows Explorer GUI. Audit ACL (SACL) You can set audit ACL (SACL) on any object in the Data Domain File System (DDFS), either through commands or through the CIFS protocol using the Windows Explorer GUI.
Setting DACL permissions using the Windows Explorer Use Explorer properties settings to select DACL permissions. Procedure 1. Right-click the file or folder and select Properties. 2. In the Properties dialog box, click the Security tab. 3. Select the group or user name, such as Administrators, from the list. The permissions appear, in this case for Administrators, Full Control. 4. Click the Advanced button, which enables you to set special permissions. 5. In the Advanced Security Settings for ACL dialog box, click the Permissions tab.
File access 215 CIFS
6. Select the permission entry in the list. 7. To view more information about a permission entry, select the entry and click Edit. 8. Select the Inherit from parent option to have the permissions of parent entries inherited by their child objects, and click OK.
Setting SACL permissions using the Windows Explorer Use Explorer properties settings to select SACL permissions. Procedure 1. Right-click the file or folder and select Properties from the menu. 2. In the Properties dialog box, click the Security tab. 3. Select the group or user name, such as Administrators, from the list, which displays its permissions, in this case, Full Control. 4. Click the Advanced button, which enables you to set special permissions. 5. In the Advanced Security Settings for ACL dialog box, click the Auditing tab. 6. Select the auditing entry in the list. 7. To view more information about special auditing entries, select the entry and click Edit. 8. Select the Inherit from parent option to have the permissions of parent entries inherited by their child objects, and click OK.
Viewing or changing the current owner security ID (owner SID) Use the Advanced Security Settings for ACL dialog box. Procedure 1. In the Advanced Security Settings for ACL dialog box, click the Owner tab. 2. To change the owner, select a name from the Change owner list, and click OK.
Controlling ID account mapping The CIFS option idmap-type controls ID account mapping behavior. This option has two values: rid (the default) and none. When the option is set to rid, the ID-to-id mapping is performed internally. When the option is set to none, all CIFS users are mapped to a local UNIX user named “cifsuser” belonging to the local UNIX group users. Consider the following information while managing this option.
l CIFS must be disabled to set this option. If CIFS is running, disable CIFS services. l The idmap-type can be set to none only when ACL support is enabled. l Whenever the idmap type is changed, a file system metadata conversion might be required for correct file access. Without any conversion, the user might not be able to access the data. To convert the metadata, consult your contracted support provider.
216 EMC Data Domain Operating System 5.7 Administration Guide CIFS
Monitoring CIFS operation Monitoring CIFS Operation topics.
Displaying CIFS status View and enable/disable CIFS status. Procedure
1. In the DD System Manager, select Protocols > CIFS.
l Status is either enabled and running, or disabled but CIFS authentication is running. To enable CIFS, see the section regarding enabling CIFS services. To disable CIFS, see the section regarding disabling CIFS services. l Connections lists the tally of open connections and open files.
Table 96 Connections Details information
Open Connections Open CIFS connections Connection Limit Maximum allowed connections Open Files Current open files Max Open Files Maximum number of open files on a Data Domain system
2. Click Connection Details to see more connection information.
Table 97 Connections Details information
Sessions Active CIFS sessions Computer IP address or computer name connected with DDR for the session User User operating the computer connected with the DDR Open Files Number of open files for each session Connection Time Connection length in minutes User Domain name of computer Mode File permissions Locks Number of locks on the file Files File location
Display CIFS configuration This section displays CIFS Configuration.
Monitoring CIFS operation 217 CIFS
Authentication configuration The information in the Authentication panel changes, depending on the type of authentication that is configured. Click the Configure link in to the left of the Authentication label in the Configuration tab. The system will navigate to the Administration > Access > Authentication page where you can configure authentication for Active Directory, Kerberos, Workgroups, and NIS. Active directory configuration Table 98 Active directory configuration information
Mode The Active Directory mode displays. Realm The configured realm displays. DDNS The status of the DDNS Server displays: either enabled or disabled. Domain Controllers The name of the configured domain controllers display or a * if all controllers are permitted. Organizational Unit The name of the configured organizational units displays. CIFS Server Name The name of the configured CIFS server displays. WINS Server Name The name of the configured WINS server displays. Short Domain Name The short domain name displays.
Workgroup configuration Table 99 Workgroup configuration authentication information
Mode The Workgroup mode displays. Workgroup Name The configured workgroup name displays. DDNS The status of the DDNS Server displays: either enabled or disabled. CIFS Server Name The name of the configured CIFS server displays. WINS Server Name The name of the configured WINS server displays.
Display shares information This section displays shares information.
218 EMC Data Domain Operating System 5.7 Administration Guide CIFS
Viewing configured shares View the list of configured shares.
Table 100 Configured shares information
Share Name The name of the share (for example, share1). Share Status The status of the share: either enabled or disabled.
Directory Path The directory path to the share (for example, /data/col1/ backup/dir1).
col1 uses the lower case letter L followed by the number 1.
Directory Path Status The status of the directory path. l To list information about a specific share, enter the share name in the Filter by Share Name text box and click Update. l Click Update to return to the default list. l To page through the list of shares, click the arrows at the bottom right of the view to page forward or backward. To skip to the beginning of the list, click | |. l Click the Items per Page drop-down arrow to change the number of share entries listed on a page. Choices are 15, 30, or 45 entries. Viewing detailed share information Display detailed information about a share by clicking a share name in the share list.
Table 101 Share information
Share Name The name of the share (for example, share1).
Directory Path The directory path to the share (for example, /data/col1/ backup/dir1).
col1 uses the lower case letter L followed by the number 1.
Directory Path Status Indicates whether the configured directory path exists on the DDR. Possible values are Path Exists or Path Does Not Exist, the later indicating an incorrect or incomplete CIFS configuration. Max Connections The maximum number of connections allowed to the share at one time. The default value is Unlimited. Comment The comment that was configured when the share was created. Share Status The status of the share: either enabled or disabled.
Display CIFS configuration 219 CIFS
l The Clients area lists the clients that are configured to access the share, along with a client tally beneath the list. l The User/Groups area lists the names and type of users or groups that are configured to access the share, along with a user or group tally beneath the list. l The Options area lists the name and value of configured options.
Displaying CIFS statistics Use the command line to display CIFS statistics. Procedure
l Enter: cifs show detailed-stats The output shows number of various SMB requests received and the time taken to process them.
Performing CIFS troubleshooting This section provides basic troubleshooting procedures.
The cifs troubleshooting commands provide detailed information about CIFS users and groups.
Displaying clients current activity Use the command line to display CIFS sessions and open files information. Procedure
l Enter: cifs show active Results
Table 102 Sessions
Computer User Open Connect Idle time files time (sec) (sec)
::ffff: ddve-25179109\sysadmin 1 92 0 10.25.132.84
Table 103 Open files
User Mode Locks File
ddve-25179109\sysadmin 1 0 C:\data\col1\backup
220 EMC Data Domain Operating System 5.7 Administration Guide CIFS
Setting the maximum open files on a connection Use the command line to set the maximum number of files that can be open concurrently. Procedure
l Enter: cifs option set max-global-open-files value. The value for the maximum global open files can be between 1 and the open files maximum limit. The maximum limit is based on the DDR system memory. For systems with greater than 12 GB, the maximum open files limit is 30,000. For systems with less than or equal to 12 GB, the maximum open files limit is 10,000.
Table 104 Connection and maximum open file limits
DDR Models Memory Connection Limit Open File Maximum Limit
DD620, DD630, DD640 8 GB 300 10,000
DD640 16 GB 600 30,000
DD640 20 GB 600 30,000
DD860 36 GB 600 30,000
DD860, DD860ArT 72 GB 600 30,000
96 GB 600 30,000
128 GB 600 30,000
256 GB 600 30,000
The system has a maximum limit of 600 CIFS connections and 250,000 open files. However, if the system runs out of open files, the number of files can be increased.
File access latencies are affected by the number of files in a directory. To the extent possible, we recommend directory sizes of less than 250,000. Larger directory sizes might experience slower responses to metadata operations such as listing the files in the directory and opening or creating a file.
Data Domain system clock When using active directory mode for CIFS access, the Data Domain System clock time can differ by no more than five minutes from that of the domain controller. The DD System Manager, Administration > Settings > Time and Date Settings tab synchronizes the clock with a time server. Because the Windows domain controller obtains the time from an external source, NTP must be configured. See the Microsoft documentation on how to configure NTP for the Windows operating system version or service pack that is running on your domain controller.
Setting the maximum open files on a connection 221 CIFS
In active directory authentication mode, the Data Domain system periodically synchronizes the clock with a Windows Active Directory Domain Controller.
Synchronizing from a Windows domain controller Use the command line on a Windows domain controller to synchronize with an NTP server.
This example is for Windows 2003 SP1; substitute your domain server for the NTP server’s name (ntpservername).
Procedure 1. On the Windows system, enter commands similar to the following:
C:\>w32tm /config /syncfromflags:manual /manualpeerlist: ntp- server-name C:\>w32tm /config /update C:\>w32tm /resync 2. After NTP is configured on the domain controller, configure the time server synchronization, as described in the section about working with time and date settings.
Synchronize from an NTP server Configure the time server synchronization, as described in the section regarding working with time and date settings.
222 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 9 NFS
This chapter includes:
l NFS overview. 224 l Managing NFS client access to the Data Domain system. 225 l Displaying NFS information. 228 l Integrating a DDR into a Kerberos domain. 229 l Add and delete KDC servers after initial configuration. 231
NFS overview Network File System (NFS) clients can have access to the system directories or MTrees on the Data Domain system. l The/backup directory is the default destination for non-MTree compressed backup server data. l The /data/col1/backup path is the root destination when using MTrees for compressed backup server data. l The /ddvar/core directory contains Data Domain System core and log files (remove old logs and core files to free space in this area).
You can also delete core files from the /ddvar or the /ddvar/ext directory if it exists.
Clients, such as backup servers that perform backup and restore operations with a Data Domain System, need access to the /backup or /data/col1/backup areas. Clients that have administrative access need to be able to access the /ddvar/core directory to retrieve core and log files. As part of the initial Data Domain system configuration, NFS clients were configured to access these areas. This chapter describes how to modify these settings and how to manage data access.
l For information about the initial system configuration, see the EMC Data Domain Operating System Initial Configuration Guide. l The nfs command manages backups and restores between NFS clients and Data Domain systems, and it displays NFS statistics and status. For complete information about the nfs command, see the EMC Data Domain Operating System Command Reference Guide. l For information about setting up third-party clients to use the Data Domain system as a server, see the related tuning guide, such as the Solaris System Tuning, which is available from the Data Domain support web site. From the Documentation > Integration Documentation page, select the vendor from the list and click OK. Select the tuning guide from the list.
HA systems and NFS HA systems are compatible with NFS. If a NFS job is in progress during a failover, the job will not need to be restarted. "/ddvar is an ext3 file system, and cannot be shared like a normal MTree-based share. The information in /ddvar will become stale when the active node fails over to the standby node because the filehandles are different on the two nodes. If /ddvar is mounted to access log files or upgrade the system, unmount and remount /ddvar if a failover has occurred since the last time /ddvar was mounted."
224 EMC Data Domain Operating System 5.7 Administration Guide NFS
Managing NFS client access to the Data Domain system The topics in this section describe how to manage NFS client access to a Data Domain System.
Enabling NFS services Enable NFS services to allow the client to access the system using the NFS protocol. Procedure
1. Select Protocols > NFS. The NFS view opens displaying the Exports tab.
Disabling NFS services Disable NFS services to prevent the client access to the system using the NFS protocol. Procedure
1. Select the Protocols > NFS tabs. The NFS view opens displaying the Exports tab.
2. Click Disable.
Creating an export You can use Data Domain System Manager’s Create button on the NFS view or use the Configuration Wizard to specify the NFS clients that can access the /backup, /data/ col1/backup,/ddvar, /ddvar/core areas, or the/ddvar/ext area if it exists. A Data Domain system supports a maximum number of 128 NFS exports.2 And 900 simultaneous connections are allowed.
You have to assign client access to each export separately and remove access from each export separately. For example, a client can be removed from /ddvar and still have access to /data/col1/backup.
If Replication is to be implemented, a single destination Data Domain system can receive backups from both CIFS clients and NFS clients as long as separate directories or MTrees are used for each. Do not mix CIFS and NFS data in the same area.
Procedure 1. Select ProtocolsNFS. The NFS view opens displaying the Exports tab.
2. May be affected by hardware limitations.
Managing NFS client access to the Data Domain system 225 NFS
2. Click Create. 3. Enter the pathname in the Directory Path text box (for example, /data/col1/ backup/dir1).
col1 uses the lower-case letter L followed by the number 1.
4. In the Clients area, select an existing client or click the + icon to create a client. The Client dialog box is displayed.
a. Enter a server name in the text box. Enter fully qualified domain names, hostnames, or IP addresses. A single asterisk (*) as a wild card indicates that all backup servers are to be used as clients.
Clients given access to the /data/col1/backup directory have access to the entire directory. A client given access to a subdirectory of /data/col1/backup has access only to that subdirectory.
l A client can be a fully-qualified domain hostname, an IPv4 or IPv6 IP address, an IPv4 address with either a netmask or prefix length, an IPv6 address with prefix length, an NIS netgroup name with the prefix @, or an asterisk (*) wildcard with a domain name, such as *.yourcompany.com. l A client added to a subdirectory under /data/col1/backup has access only to that subdirectory. l Enter an asterisk (*) as the client list to give access to all clients on the network.
b. Select the checkboxes of the NFS options for the client. General:
l Read-only permission (ro). l Allow connections from ports below 1024 (secure) (default). l Log NFS request arguments (log). This option may impact performance.
l Map requests from UID (user identifier) or GID (group identifier) 0 to the anonymous UID/GID (root _squash). l Map all user requests to the anonymous UID/GID (all _squash). l Use Default Anonymous UID/GID.
Allowed Kerberos Authentication Modes:
l Unauthenticated connections (sec=sys). Select to not use authentication. l Authenticated Connections (sec=krb5).
Integrity and Privacy are not supported.
226 EMC Data Domain Operating System 5.7 Administration Guide NFS
c. Click OK. 5. Click OK to create the export.
Modifying an export Change the directory path, domain name, and other options using the GUI. Procedure
1. SelectProtocols > NFS. The NFS view opens displaying the Exports tab.
2. Click the checkbox of an export in the NFS Exports table. 3. Click Modify. 4. Modify the pathname in the Directory Path text box. 5. In the Clients area, select another client and click the pencil icon (modify), or click the + icon to create a client. a. Enter a server name in the Client text box. Enter fully qualified domain names, hostnames, or IP addresses. A single asterisk (*) as a wild card indicates that all backup servers are to be used as clients.
Clients given access to the /data/col1/backup directory have access to the entire directory. A client given access to a subdirectory of /data/col1/backup has access only to that subdirectory.
l A client can be a fully-qualified domain hostname, an IPv4 or IPv6 IP address, an IPv4 address with either a netmask or prefix length, an IPv6 address with prefix length, an NIS netgroup name with the prefix @, or an asterisk (*) wildcard with a domain name, such as *.yourcompany.com. A client added to a subdirectory under /data/col1/backup has access only to that subdirectory. l Enter an asterisk (*) as the client list to give access to all clients on the network.
b. Select the checkboxes of the NFS options for the client. General:
l Read-only permission (ro). l Allow connections from ports below 1024 (secure) (default). l Log NFS request arguments (log). This option may impact performance.
l Map requests from UID (user identifier) or GID (group identifier) 0 to the anonymous UID/GID (root _squash). l Map all user requests to the anonymous UID/GID (all _squash). l Use Default Anonymous UID/GID.
Allowed Kerberos Authentication Modes:
Modifying an export 227 NFS
l Unauthenticated connections (sec=sys). Select to not use authentication. l Authenticated Connections (sec=krb5).
Integrity and Privacy are not supported.
c. Click OK. 6. Click OK to modify the export.
Creating an export from an existing export Create an export from an existing export and then modify it as needed. Procedure 1. In the NFS Exports tab, click the checkbox of the export you wish to use as the source. 2. Click Create From. 3. Modify the export information, as described in section about modifying an export.
Deleting an export Delete an export from the NFS Exports tab. Procedure 1. In the NFS Exports tab, click the checkbox of the export you wish to delete. 2. Click Delete. 3. Click OK and Close to delete the export.
Displaying NFS information The topics in this section describe how to use the DD System Manager to monitor NFS client status and NFS configuration.
Viewing NFS status Display whether NFS is active and Kerberos is enabled. Procedure
l Click Protocols > NFS. The top panel shows the operational status of NFS; for example, whether NFS is currently active and running, and whether Kerberos mode is enabled.
Click Configure to view the Administration > Access > Authentication tab where you can configure Kerberos authentication.
228 EMC Data Domain Operating System 5.7 Administration Guide NFS
Viewing NFS exports See the list of clients allowed to access the Data Domain System. Procedure 1. Select ProtocolsNFS. The Exports view shows a table of NFS exports that are configured for Data Domain System and the mount path, status, and NFS options for each export.
2. Click an export in the table to populate the Detailed Information area, below the Exports table. In addition to the export’s directory path, configured options, and status, the system displays a list of clients. Use the Filter By text box to sort by mount path. Click Update for the system to refresh the table and use the filters supplied. Click Reset for the system to clear the Path and Client filters.
Viewing active NFS clients Display all clients that have been connected in the past 15 minutes and their mount path. Procedure
l Select the Protocols > NFS > Active Clients tab. The Active Clients view displays, showing all clients that have been connected in the past 15 minutes and their mount path. Use the Filter By text boxes to sort by mount path and client name. Click Update for the system to refresh the table and use the filters supplied. Click Reset for the system to clear the Path and Client filters.
Integrating a DDR into a Kerberos domain Set the domain name, the host name, and the DNS server for the DDR. Enable the DDR to use the authentication server as a Key Distribution Center (for UNIX) and as a Distribution Center (for Windows Active Directory).
The examples provided in this description are specific to the operating system (OS) used to develop this exercise. You must use commands specific to your OS.
For UNIX Kerberos mode, a keytab file must be transferred from the Key Distribution Center (KDC) server, where it is generated, to the DDR. If you are using more than one DDR, each DDR requires a separate keytab file. The keytab file contains a shared secret between the KDC server and the DDR.
Viewing NFS exports 229 NFS
When using a UNIX KDC, the DNS server does not have to be the KDC server, it can be a separate server.
Procedure 1. Set the host name and the domain name for the DDR, using DDR commands.
net set hostname
The host name is the name of the DDR.
2. Configure NFS principal (node) for the DDR on the Key Distribution Center (KDC). Example: addprinc nfs/hostname@realm
Hostname is the name for the DDR.
3. Verify that there are nfs entries added as principals on the KDC. Example: listprincs nfs/hostname@realm
4. Add the DDR principal into a keytab file. Example: ktadd nfs/hostname@realm
5. Verify that there is an nfs keytab file configured on the KDC. Example: klist -k
The is the keytab file used to configure keys in a previous step.
6. Copy the keytab file from the location where the keys for NFS DDR are generated to the DDR in the /ddvar/ directory.
Table 105 Keytab destination
Copy file from: Copy file to:
(The keytab file configured in a previous /ddvar/ step.)
7. Set the realm on the DDR, using the following DDR command:
230 EMC Data Domain Operating System 5.7 Administration Guide NFS
authentication kerberos set realm kdc-type kdcs 8. When the kdc-type is UNIX, import the keytab file from /ddvar/ to /ddr/etc/, where the Kerberos configuration file expects it. Use the following DDR command to copy the file: authentication kerberos keytab import NOTICE
This step is required only when the kdc-type is UNIX.
Kerberos setup is now complete. 9. To add a NFS mount point to use Kerberos, use the nfs add command. See the EMC Data Domain Operating System Command Reference Guide for more information. 10. Add host, NFS and relevant user principals for each NFS client on the Key Distribution Center (KDC). Example: listprincs host/hostname@realm nfs/hostname@realm root/hostname@realm
11. For each NFS client, import all its principals into a keytab file on the client. Example: ktadd -k host/hostname@realm
ktadd -k nfs/hostname@realm
Add and delete KDC servers after initial configuration After you have integrated a DDR into a Kerberos domain, and thereby enabled the DDR to use the authentication server as a Key Distribution Center (for UNIX) and as a Distribution Center (for Windows Active Directory), you can use the following procedure to add or delete KDC servers. Procedure 1. Join the DDR to a Windows Active Directory (AD) server or a UNIX Key Distribution Center (KDC). authentication kerberos set realm kdc-type
Example: authentication kerberos set realm krb5.test kdc-type unix kdcs nfskrb-kdc.krb5.test This command joins the system to the krb5.test realm and enables Kerberos authentication for NFS clients.
A keytab generated on this KDC must exist on the DDR to authenticate using Kerberos. 2. Verify the Kerberos authentication configuration. authentication kerberos show config
Add and delete KDC servers after initial configuration 231 NFS
Home Realm: krb5.test KDC List: nfskrb-kdc.krb5.test KDC Type: unix
3. Add a second KDC server.
authentication kerberos set realm kdc-type
Example: authentication kerberos set realm krb5.test kdc-type unix kdcs ostqa-sparc2.krb5.test nfskrb-kdc.krb5.test
A keytab generated on this KDC must exist on the DDR to authenticate using Kerberos.
4. Verify that two KDC servers are added.
authentication kerberos show config Home Realm: krb5.test KDC List: ostqa-sparc2.krb5.test, nfskrb-kdc.krb5.test KDC Type: unix
5. Display the value for the Kerberos configuration key.
reg show config.keberos config.kerberos.home_realm = krb5.test config.kerberos.home_realm.kdc1 = ostqa-sparc2.krb5.test config.kerberos.home_realm.kdc2 = nfskrb-kdc.krb5.test config.kerberos.kdc_count = 2 config.kerberos.kdc_type = unix 6. Delete a KDC server.
Delete a KDC server by using the authentication kerberos set realm kdc-type command without listing the KDC server that you want to delete. For example, if the existing KDC servers are kdc1, kdc2, and kdc3, and you want to remove kdc2 from the realm, you could use the following example:
authentication kerberos set realm kdc-type kdcs kdc1,kdc3
232 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 10 Storage Migration
This chapter includes:
l Storage migration overview. 234 l Migration planning considerations. 234 l Viewing migration status. 235 l Evaluating migration readiness. 236 l Migrating storage using DD System Manager. 236 l Storage migration dialog descriptions. 237 l Migrating storage using the CLI. 240 l CLI storage migration example. 241
Storage Migration 233 Storage Migration
Storage migration overview Storage migration supports the replacement of existing storage enclosures with new enclosures that may offer higher performance, higher capacity, and a smaller footprint. After new enclosures are installed, you can migrate the data from the older enclosures to the new enclosures while the system continues to support other processes such as data access, expansion, cleaning, and replication. The storage migration does require system resources, but you can control this with throttle settings that give the migration a relatively higher or lower priority. You can also suspend a migration to make more resources available to other processes, then resume the migration when resource demand is lower. During the migration, the system uses data on the source and destination enclosures. New data is written to the new enclosures. Non-migrated data is updated on the source enclosures, and migrated data is updated on the destination enclosures. If the migration is interrupted, the migration can resume migrating blocks that have not been marked as migrated. During the migration, each block of data is copied and verified, the source block is freed and marked as migrated, and the system index is updated to use the new location. New data that was destined to land in the source block will now be redirected to destination block. All new data block allocations that would have been allocated from source are allocated from the destination. The Migration copy process is done at the shelf level, not the logical data level, so all disk sectors on the source shelf are accessed and copied over regardless of whether there is data on them. Therefore, the Storage Migration Utility cannot be used to shrink a logical data footprint.
Because the data set is divided between the source and destination enclosures during migration, you cannot halt a migration and resume use of only the source enclosures. Once started, the migration must complete. If a failure, such as a faulty disk drive, interrupts the migration, address the issue and resume the migration.
Depending on the amount of data to migrate and the throttle settings selected, a storage migration can take days or weeks. When all data is migrated, the finalize process, which must be manually initiated using the storage migration finalize command, restarts the filesystem. During the restart, the source enclosures are removed from the system configuration and the destination enclosures become part of the filesystem. When the finalize process is complete, the source enclosures can be removed from the system.
Migration planning considerations Consider the following guidelines before starting a storage migration.
l Storage migration requires a single-use license and operates on system models supported by DD OS version 5.7 or later.
Multiple storage migration needs multiple licenses.
l The destination enclosures must:
234 EMC Data Domain Operating System 5.7 Administration Guide Storage Migration
n Be new, unassigned shelves. n Be supported on the DD system model. n Contain at least as much usable capacity as the enclosures they are replacing.
Migration between enclosures with identical raw capacity can fail if the usable capacity on the destination is less than that on the source. For example, enclosures with higher capacity disk drives can have a proportionately larger usable capacity than that for enclosures with smaller capacity disk drives. In this situation, a migration to the enclosures with smaller drives might fail if data storage in the source enclosures is approaching full capacity.
l Data migration is not supported for disks in the system controller. l Although DD OS can be upgraded while a migration is taking place, we recommend that you do not upgrade DD OS during the migration finalize phase. l Storage migration cannot start when the file system is disabled or while a DD OS upgrade is in progress, another migration is in progress, or a RAID reconstruction is in progress. l All specified source enclosures must be in the same tier (active or archive). l There can be only one disk group in each source enclosure, and all disks in the disk group must be installed in within the same enclosure. l All disks in each destination enclosure must be of the same type (for example, all SATA or all SAS). l After migration begins, the destination enclosures cannot be removed. l Source enclosures cannot be removed until migration is complete and finalized. l The storage migration duration depends on the system resources (which differ for different system models), the availability of system resources, and the data quantity to migrate. Storage migration can take days or weeks to complete.
Viewing migration status DD SM provides two ways to view storage migration status. Procedure 1. Select Hardware > Storage. In the Storage area, review the Storage Migration Status line. If the status is Not Licensed, you must add a license before using any storage migration features. If the storage migration license is installed, the status can be one of the following: None, Starting, Migrating, Paused by User, Paused by System, Copy Completed - Pending Finalization, Finalizing, Failed during Copy, or Failed during Finalize. 2. If a storage migration is in progress, click View Storage Migration to view the progress dialogs.
The migration status shows the percentage of blocks transferred. In a system with many free blocks, the free blocks are not migrated, but they are included in the progress indication. In this situation, the progress indication will climb quickly and then slow when the data migration starts.
Viewing migration status 235 Storage Migration
3. When a storage migration is in progress, you can also view the status by selecting Health > Jobs.
Evaluating migration readiness You can use the system to evaluate storage migration readiness without committing to start the migration. Procedure 1. Install the destination enclosures using the instructions in the product installation guides. 2. Select Administration > Licenses and verify that the storage migration license is installed. 3. If the storage migration license is not installed, click Add Licenses and add the license. 4. Select Hardware > Storage, then click Migrate Data.
5. In the Select a Task dialog, select Estimate, then click Next. 6. In the Select Existing Enclosures dialog, use the checkboxes to select each of the source enclosures for the storage migration, then click Next. 7. In the Select New Enclosures dialog, use the checkboxes to select each of the destination enclosures for the storage migration, then click Next. The Add Licenses button allows you to add storage licenses for the new enclosures as needed, without interrupting the current task.
8. In the Review Migration Plan dialog, review the estimated migration schedule, then click Next. 9. Review the precheck results in the Verify Migration Preconditions dialog, then click Close. Results If any of the precheck tests fail, resolve the issue before you start the migration.
Migrating storage using DD System Manager The storage migration process evaluates system readiness, prompts you to confirm that you want to start the migration, migrates the data, and then prompts you to finalize the process. Procedure 1. Install the destination enclosures using the instructions in the product installation guides. 2. Select Administration > Licenses and verify that the storage migration license is installed. 3. If the storage migration license is not installed, click Add Licenses and add the license. 4. Select Hardware > Storage, then click Migrate Data.
5. In the Select a Task dialog, select Migrate, then click Next.
236 EMC Data Domain Operating System 5.7 Administration Guide Storage Migration
6. In the Select Existing Enclosures dialog, use the checkboxes to select each of the source enclosures for the storage migration, then click Next. 7. In the Select New Enclosures dialog, use the checkboxes to select each of the destination enclosures for the storage migration, then click Next. The Add Licenses button allows you to add storage licenses for the new enclosures as needed, without interrupting the current task.
8. In the Review Migration Plan dialog, review the estimated migration schedule, then click Start. 9. In the Start Migration dialog, click Start. The Migrate dialog appears and updates during the three phases of the migration: Starting Migration, Migration in Progress, and Copy Complete.
10. When the Migrate dialog title displays Copy Complete and a filesystem restart is acceptable, click Finalize.
This task restarts the filesystem and typically takes 10 to 15 minutes. The system is unavailable during this time.
Results When the migration finalize task is complete, the system is using the destination enclosures and the source enclosures can be removed.
Storage migration dialog descriptions The DD System Manager dialog descriptions provide additional information on storage migration. This information is also available by clicking the help icon in the dialogs.
Select a Task dialog The configuration in this dialog determines whether the system will evaluate storage migration readiness and stop, or evaluate readiness and begin storage migration. Select Estimate to evaluate system readiness and stop. Select Migrate to start migration after the system evaluation. Between the system evaluation and the start of the migration, a dialog prompts you to confirm or cancel the storage migration.
Select Existing Enclosures dialog The configuration in this dialog selects either the active or the retention tier and the source enclosures for the migration. If the Extended Retention feature is installed, use the list box to select either the Active Tier or Retention Tier. The list box does not appear when Extended Retention is not installed. The Existing Enclosures list displays the enclosures that are eligible for storage migration. Select the checkbox for each of the enclosures to migrate. Click Next when you are ready to continue.
Storage migration dialog descriptions 237 Storage Migration
Select New Enclosures dialog The configuration in this dialog selects the destination enclosures for the migration. This dialog also displays the storage license status and an Add Licenses button. The Available Enclosures list displays the enclosures that are eligible destinations for storage migration. Select the checkbox for each of the desired destination enclosures. The license status bar represents all of the storage licenses installed on the system. The green portion represents licenses that are in use, and the and clear portion represents the licensed storage capacity available for destination enclosures. If you need to install additional licenses to support the selected destination controllers, click Add Licenses. Click Next when you are ready to continue.
Review Migration Plan dialog This dialog presents an estimate of the storage migration duration, organized according to the three stages of storage migration. Stage 1 of the storage migration runs a series of tests to verify that the system is ready for the migration. The test results appear in the Verify Migration Preconditions dialog. During Stage 2, the data is copied from the source enclosures to the destination enclosures. When a large amount of data is present, the copy can take days or weeks to complete because the copy takes place in the background, while the system continues to serve backup clients. A setting in the Migration in Progress dialog allows you to change the migration priority, which can speed up or slow down the migration. Stage 3, which is manually initiated from the Copy Complete dialog, updates the system configuration to use the destination enclosures and removes the configuration for the source controllers. During this stage, the file system is restarted and the system is unavailable to backup clients.
Verify Migration Preconditions dialog This dialog displays the results of the tests that execute before the migration starts. The following list shows the test sequence and provides additional information on each of the tests. P1. This system's platform is supported. Older DD system models do not support storage migration. P2. A storage migration license is available. A storage migration license is required. P3. No other migration is currently running. A previous storage migration must complete before you can start another. P4. The current migration request is the same as the interrupted migration request. Resume and complete the interrupted migration. P5. Check the disk group layout on the existing enclosures. Storage migration requires that each source enclosure contain only one disk group, and all the disks in the group must be in that enclosure.
238 EMC Data Domain Operating System 5.7 Administration Guide Storage Migration
P6. Verify the final system capacity. The total system capacity after migration and the removal of the source enclosures must not exceed the capacity supported by the DD system model. P7. Verify the replacement enclosures' capacity. The usable capacity of the destination enclosures must be greater than that of the source enclosures. P8. Source enclosures are in the same active tier or retention unit. The system supports storage migration from either the active tier or the retention tier. It does not support migration of data from both tiers at the same time. P9. Source enclosures are not part of the head unit. Although the system controller is listed as an enclosure in the CLI, storage migration does not support migration from disks installed in the system controller. P10. Replacement enclosures are addable to storage. All disks in each destination enclosure must be of the same type (for example, all SATA or all SAS). P11. No RAID reconstruction is occurring in the source controllers. Storage migration cannot start while a RAID reconstruction is in progress.
Migration progress dialogs This series of dialogs presents the storage migration status and the controls that apply at each stage. Migrate - Starting Migration During the first stage, the progress is shown on the progress bar and no controls are available. Migrate - Migration in Progress During the second stage, data is copied from the source enclosures to the destination enclosures and the progress is shown on the progress bar. Because the data copy can take days or weeks to complete, controls are provided so that you can manage the resources used during migration and suspend migration when resources are needed for other processes. You can click Pause to suspend the migration and later click Resume to continue the migration. The Low, Medium, and High buttons define throttle settings for storage migration resource demands. A low throttle setting gives storage migration a lower resource priority, which results in a slower migration and requires fewer system resources. Conversely, A high throttle setting gives storage migration a higher resource priority, which results in a faster migration and requires more system resources. The medium setting selects an intermediate priority. You do not have to leave this dialog open for the duration of the migration. To check the status of the migration after closing this dialog, select Hardware > Storage and view the migration status. To return to this dialog from the Hardware/Storage page, click Manage Migration. The migration progress can also be viewed by selecting Health > Jobs. Migrate - Copy Complete When the copy is complete, the migration process waits for you to click Finalize. During this final stage, , which takes 10 to 15 minutes, the filesystem is restarted and the
Migration progress dialogs 239 Storage Migration
system is not available. It is a good practice to start this stage during a maintenance window or a period of low system activity.
Migrating storage using the CLI A migration simply requires moving all of the allocated blocks from the blocksets formatted over source DGs (e.g., source blocksets) to the blocksets formatted over destination DGs (e.g., destination blocksets). Once all of the allocated blocks have been moved from the source blocksets, those blocksets can be removed from DDFS, their disks can be removed from their storage tier, and the physical disks and enclosures can be removed from the DDR.
The preparation of new enclosures for storage migration is managed by the storage migration process. Do not prepare destination enclosures as you would for an enclosure addition. For example, use of the filesys expand command is appropriate for an enclosure addition, but this command prevents enclosures from being used as storage migration destinations.
Procedure 1. Install the destination enclosures using the instructions in the product installation guides. 2. Check to see if the storage migration feature license is installed. # license show 3. If the storage migration license is not installed, install it now. # license add AAAA-BBBB-CCCC-DDDD-EEEE-FFFF-GGGG-HHHH 4. View the disk states for the source and destination disks. # disk show state The source disks should be in the active state, and the destination disks should be in the unknown state. 5. Run the storage migration precheck command to determine if the system is ready for the migration. # storage migration precheck source-enclosures 7:2 destination- enclosures 7:4 6. View the migration throttle setting. storage migration option show throttle
7. When the system is ready, begin the storage migration. # storage migration start source-enclosures 7:2 destination- enclosures 7:4
8. Review the migration status as needed. # storage migration status
9. View the disk states for the source and destination disks. # disk show state During the migration, the source disks should be in the migrating state, and the destination disks should be in the destination state.
240 EMC Data Domain Operating System 5.7 Administration Guide Storage Migration
10. When the migration is complete, update the configuration to use the destination enclosures.
This task restarts the filesystem and typically takes 10 to 15 minutes. The system is unavailable during this time.
storage migration finalize
11. If you want to remove all data from each of the source enclosures, remove the data now. storage sanitize start enclosure [:
The storage sanitize command does not produce a certified data erasure. EMC offers certified data erasure as a service. For more information, contact your EMC representative.
12. View the disk states for the source and destination disks. # disk show state After the migration, the source disks should be in the unknown state, and the destination disks should be in the active state. Results When the migration finalize task is complete, the system is using the destination storage and the source storage can be removed.
CLI storage migration example
CLI storage migration example 241 Storage Migration
Figure 8 Output: license show
Figure 9 Output: license add
Figure 10 Output: disk show state
242 EMC Data Domain Operating System 5.7 Administration Guide Storage Migration
Figure 11 Output: storage migration precheck
Figure 12 Output: storage migration show history
CLI storage migration example 243 Storage Migration
Figure 13 Output: storage migration start, part 1
Figure 14 Output: storage migration start, part 2
Figure 15 Output: storage migration status
244 EMC Data Domain Operating System 5.7 Administration Guide Storage Migration
Figure 16 Output: disk show state, migration in progress
Figure 17 Output: storage migration finalize
CLI storage migration example 245 Storage Migration
Figure 18 Output: disk show state, migration complete
Currently storage migration is only supported on Active node. Storage migration is not supported on passive node of the HA cluster.
246 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 11 SCSI Target
This chapter includes:
l SCSI Target overview. 248 l Fibre Channel view. 249 l Differences in FC link monitoring among DD OS versions. 259
SCSI Target 247 SCSI Target
SCSI Target overview SCSI (Small Computer System Interface) Target is a unified management daemon for all SCSI services and transports. SCSI Target supports VTL (Virtual Tape Library), DD Boost over FC (Fibre Channel), and vDisk/ProtectPoint Block Services, as well as anything that has a target LUN (logical unit number) on a DD system. SCSI Target Services and Transports The SCSI Target daemon starts when FC ports are present or VTL is licensed. It provides unified management for all SCSI Target services and transports.
l A service is anything that has a target LUN on a DD system that uses SCSI Target commands, such as VTL (tape drives and changers), DD Boost over FC (processor devices), or vDisk (Virtual Disk Device). l A transport enables devices to become visible to initiators. l An initiator is a backup client that connects to a system to read and write data using the FC protocol. A specific initiator can support DD Boost over FC, vDisk, or VTL, but not all three. l Devices are visible on a SAN (storage area network) through physical ports. Host initiators communicate with the DD system through the SAN. l Access groups manage access between devices and initiators. l An endpoint is the logical target on a DD system to which an initiator connects. You can disable, enable, and rename endpoints. To delete endpoints, the associated transport hardware must no longer exist. Endpoints are automatically discovered and created when a new transport connection occurs. Endpoints have the following attributes: port topology, FCP2-RETRY status, WWPN, and WWNN. l NPIV (N_port ID Virtualization) is an FC feature that lets multiple endpoints share a single physical port. NPIV eases hardware requirements and provides failover capabilities. Note the following exceptions:
l VTL, vDisk/ProtectPoint Block Services, and DD Boost over FC can run on the same system, but they cannot share the same HBA (host bus adapter) ports. l DD Boost can service both FC and IP clients simultaneously; however, both transports cannot share the same initiator. l Only one initiator should be present per access group. Each access group is assigned a type (VTL, vDisk/ProtectPoint Block Services, or DD Boost over FC). SCSI Target Architectures - Supported and Unsupported SCSI Target supports the following architectures:
l VTL plus DD Boost over FC from different initiators: Two different initiators (on the same or different clients) may access a DD system using VTL and DD Boost over FC, through the same or different DD system target endpoints. l VTL plus DD Boost over FC from one initiator to two different DD systems: A single initiator may access two different DD systems using any service. SCSI Target does not support the following architecture:
l VTL plus DD Boost over FC from one initiator to the same DD system: A single initiator may not access the same DD system through different services. For More Information about DD Boost and the scscitarget Command (CLI)
248 EMC Data Domain Operating System 5.7 Administration Guide SCSI Target
For more information about using DD Boost through the DD System Manager, see the related chapter in this book. For other types of information about DD Boost, see the EMC Data Domain Boost for OpenStorage Administration Guide. This chapter focuses on using SCSI Target through the DD System Manager. After you have become familiar with basic tasks, the scscitarget command in the EMC Data Domain Operating System Command Reference Guide provides more advanced management tasks. When there is heavy VTL traffic, avoid running the scsitarget group use command, which switches the in-use endpoint lists for one or more SCSI Target or vdisk devices in a group between primary and secondary endpoint lists.
Fibre Channel view The Fibre Channel view displays the current status of whether Fibre Channel and/or NPIV is enabled. It also displays two tabs: Resources and Access Groups. Resources include ports, endpoints, and initiators. An access group holds a collection of initiator WWPNs (worldwide port names) or aliases and the drives and changers they are allowed to access.
Enabling NPIV NPIV (N_Port ID Virtualization), is a Fibre Channel feature in which multiple endpoints can share a single physical port. NPIV eases hardware requirements and provides endpoint failover/failback capabilities. NPIV is not configured by default; you must enable it.
NPIV is enabled by default in HA configuration.
NPIV provides simplified multiple-system consolidation: l NPIV is an ANSI T11 standard that allows a single HBA physical port to register with a Fibre Channel fabic using multiple WWPNs l The virtual and physical ports have the same port properties and behave exactly the same. l There may be m:1 relationships between the endpoints and the port, that is, multiple endpoints can share the same physical port. Specifically, enabling NPIV enables the following features: l Multiple endpoints are allowed per physical port, each using a virtual (NPIV) port. The base port is a placeholder for the physical port and is not associated with an endpoint. l Endpoint failover/failback is automatically enabled when using NPIV.
After NPIV is enabled, the "Secondary System Address" must be specified at each of the endpoints. If not, endpoint failover will not occur.
l Multiple DD systems can be consolidated into a single DD system, however, the number of HBAs remains the same on the single DD system. l The endpoint failover is triggered when FC-SSM detects when a port goes from online to offline. In the case where the physical port is offline before scsitarget is enabled and the port is still offline after scsitarget is enabled, a endpoint failover is not possible because FC-SSM does not generate a port offline event. If the port comes
Fibre Channel view 249 SCSI Target
back online and auto-failback is enabled, any failed over endpoints that use that port as a primary port will fail-back to the primary port. The Data Domain HA features requires NPIV to move WWNs between the nodes of an HA pair during the failover process.
Before enabling NPIV, the following conditions must be met:
l The DD system must be running DD OS 5.7. l All ports must be connected to 4Gb, 8Gb, and 16 Gb Fibre Channel HBA and SLIC. l The DD system ID must be valid, that is, it must not be 0. In addition, port topologies and port names will be reviewed and may prevent NPIV from being enabled: l NPIV is allowed if the topology for all ports is loop-preferred. l NPIV is allowed if the topology for some of the ports is loop-preferred; however, NPIV must be disabled for ports that are loop-only, or you must reconfigure the topology to loop-preferred for proper functionality. l NPIV is not allowed if none of the ports has a topology of loop-preferred. l If port names are present in access groups, the port names are replaced with their associated endpoint names.
Procedure 1. Select Hardware > Fibre Channel. 2. Next to NPIV: Disabled, select Enable. 3. In the Enable NPIV dialog, you will be warned that all Fibre Channel ports must be disabled before NPIV can be enabled. If you are sure that you want to do this, select Yes.
CLI Equivalent a. Make sure (global) NPIV is enabled. # scsitarget transport option show npiv SCSI Target Transport Options Option Value ------npiv disabled ------b. If NPIV is disabled, then enable it. You must first disable all ports. # scsitarget port disable all All ports successfully disabled. # scsitarget transport option set npiv enabled Enabling FiberChannel NPIV mode may require SAN zoning to be changed to configure both base port and NPIV WWPNs. Any FiberChannel port names used in the access groups will be converted to their corresponding endpoint names in order to prevent ambiguity. Do you want to continue? (yes|no) [no]: c. Re-enable the disabled ports. # scsitarget port enable all All ports successfully enabled. d. Make sure the physical ports have an NPIV setting of “auto”. # scsitarget port show detailed 0a System Address: 0a
250 EMC Data Domain Operating System 5.7 Administration Guide SCSI Target
Enabled: Yes Status: Online Transport: FibreChannel Operational Status: Normal FC NPIV: Enabled (auto) . . . e. Create a new endpoint using the primary and secondary ports you have selected. # scsitarget endpoint add test0a0b system-address 0a primary-system-address 0a secondary-system-address 0b Note that the endpoint is disabled by default, so enable it. # scsitarget endpoint enable test0a0b Then display the endpoint information. # scsitarget endpoint show detailed test0a0b Endpoint: test0a0b Current System Address: 0b Primary System Address: 0a Secondary System Address: 0b Enabled: Yes Status: Online Transport: FibreChannel FC WWNN: 50:02:18:80:08:a0:00:91 FC WWPN: 50:02:18:84:08:b6:00:91 f. Zone a host system to the auto-generated WWPN of the newly created endpoint. g. Create a VTL, vDisk, or DD Boost over Fibre Channel (DFC) device, and make this device available on the host system. h. Ensure that the DD device chosen can be accessed on the host (read and/or written). i. Test the endpoint failover by using the “secondary” option to move the endpoint to the SSA (secondary system address). # scsitarget endpoint use test0a0b secondary j. Ensure that the DD device chosen can still be accessed on the host (read and/or written). Test the failback by using the “primary” option to move the endpoint back to the PSA (primary system address). # scsitarget endpoint use test0a0b primary k. Ensure that the DD device chosen can still be accessed on the host (read and/or written).
Disabling NPIV Before you can disable NPIV, you must not have any ports with multiple endpoints.
NPIV is required for HA configuration. It is enabled by default and cannot be disabled.
1. Select Hardware > Fibre Channel. 2. Next to NPIV: Enabled, select Disable. 3. In the Disable NPIV dialog, review any messages about correcting the configuration, and when ready, select OK.
Disabling NPIV 251 SCSI Target
Resources tab The Hardware > Fibre Channel > Resources tab displays information about ports, endpoints, and initiators.
Table 106 Ports
System Address System address for port WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node Enabled Port operational status; either Enabled or Disabled. NPIV NPIV status; either Enabled or Disabled. Link Status Link status: either Online or Offline; that is, whether or not the port is up and capable of handling traffic. Operation Status Operation status: either Normal or Marginal. # of Endpoints Number of endpoints associated with this port.
Table 107 Endpoints
Name Name of endpoint. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node System Address System address of endpoint. Enabled Port operational state; either Enabled or Disabled. Link Status Either Online or Offline; that is, whether or not the port is up and capable of handling traffic.
Table 108 Initiators
Name Name of initiator. Service Service support by the initiator, which is either VTL, DD Boost, or vDisk.
252 EMC Data Domain Operating System 5.7 Administration Guide SCSI Target
Table 108 Initiators (continued)
WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. Vendor Name Initiator's model. Online Endpoints Endpoints seen by this initiator. Displays none or offline if the initiator is not available.
Configuring a port Ports are discovered, and a single endpoint is automatically created for each port, at startup. The properties of the base port depend on whether NPIV is enabled:
l In non-NPIV mode, ports use the same properties as the endpoint, that is, the WWPN for the base port and the endpoint are the same. l In NPIV mode, the base port properties are derived from default values, that is, a new WWPN is generated for the base port and is preserved to allow consistent switching between NPIV modes. Also, NPIV mode provides the ability to support multiple endpoints per port. Procedure
1. Select Hardware > Fibre Channel > Resources. 2. Under Ports, select an port, and then select Modify (pencil). 3. In the Configure Port dialog, select whether to automatically enable or disable NPIV for this port. 4. For Topology, select Loop Preferred, Loop Only, Point to Point, or Default. 5. For Speed, select 1, 2, 4 or 8 Gbps, or auto. 6. Select OK.
Enabling a port Ports must be enabled before they can be used. Procedure
1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Ports > Enable. If all ports are already enabled, a message to that effect is displayed. 3. In the Enable Ports dialog, select one or more ports from the list, and select Next. 4. After the confirmation, select Next to complete the task.
Resources tab 253 SCSI Target
Disabling a port You can simply disable a port (or ports), or you can chose to failover all endpoints on the port (or ports) to another port. Procedure 1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Ports > Disable. 3. In the Disable Ports dialog, select one or more ports from the list, and select Next. 4. In the confirmation dialog, you can continue with simply disabling the port, or you can chose to failover all endpoints on the ports to another port.
Adding an endpoint An endpoint is a virtual object that is mapped to a underlying virtual port. In non-NPIV mode (not available on HA configuration), only a single endpoint is allowed per physical port, and the base port is used to configure that endpoint to the fabric. When NPIV is enabled, multiple endpoints are allowed per physical port, each using a virtual (NPIV) port, and endpoint failover/failback is enabled.
Non-NPIV mode is not available on HA configurations. NPIV is enabled by default and cannot be disabled.
In NPIV mode, endpoints:
l have a primary system address. l may have zero or more secondary system addresses. l are all candidates for failover to an alternate system address on failure of a port; however, failover to a marginal port is not supported. l may be failed back to use their primary port when the port comes back up online.
When using NPIV, it is recommended that you use only one protocol (that is, VTL Fibre Channel, DD Boost-over-Fibre Channel, or vDisk Fibre Channel) per endpoint. For failover configurations, secondary endpoints should also be configured to have the same protocol as the primary.
Procedure 1. Select Hardware > Fibre Channel > Resources. 2. Under Endpoints, select Add (+ sign). 3. In the Add Endpoint dialog, enter a Name for the endpoint (from 1 to 128 characters). The field cannot be empty or be the word “all,” and cannot contain the characters asterisk (*), question mark (?), front or back slashes (/, \), or right or left parentheses [(,)]. 4. For Endpoint Status, select Enabled or Disabled. 5. If NPIV is enabled, for Primary system address, select from the drop-down list. The primary system address must be different from any secondary system address.
254 EMC Data Domain Operating System 5.7 Administration Guide SCSI Target
6. If NPIV is enabled, for Fails over to secondary system addresses, check the appropriate box next to the secondary system address. 7. Select OK.
Configuring an endpoint After you have added an endpoint, you can modify it using the Configure Endpoint dialog.
When using NPIV, it is recommended that you use only one protocol (that is, VTL Fibre Channel, DD Boost-over-Fibre Channel, or vDisk Fibre Channel) per endpoint. For failover configurations, secondary endpoints should also be configured to have the same protocol as the primary.
1. Select Hardware > Fibre Channel > Resources. 2. Under Endpoints, select an endpoint, and then select Modify (pencil). 3. In the Configure Endpoint dialog, enter a Name for the endpoint (from 1 to 128 characters). The field cannot be empty or be the word “all,” and cannot contain the characters asterisk (*), question mark (?), front or back slashes (/, \), or right or left parentheses [(,)]. 4. For Endpoint Status, select Enabled or Disabled. 5. For Primary system address, select from the drop-down list. The primary system address must be different from any secondary system address. 6. For Fails over to secondary system addresses, check the appropriate box next to the secondary system address. 7. Select OK.
Modifying an endpoint's system address You can modify the active system address for a SCSI Target endpoint using the scsitarget endpoint modify command option. This is useful if the endpoint is associated with a system address that no longer exists, for example after a controller upgrade or when a controller HBA (host bus adapter) has been moved. When the system address for an endpoint is modified, all properties of the endpoint, including WWPN and WWNN (worldwide port and node names, respectively), if any, are preserved and are used with the new system address. In the following example, endpoint ep-1 was assigned to system address 5a, but this system address is no longer valid. A new controller HBA was added at system address 10a. The SCSI Target subsystem automatically created a new endpoint, ep-new, for the newly discovered system address. Because only a single endpoint can be associated with a given system address, ep-new must be deleted, and then ep-1 must be assigned to system address 10a.
It may take some time for the modified endpoint to come online, depending on the SAN environment, since the WWPN and WWNN have moved to a different system address. You may also need to update SAN zoning to reflect the new configuration.
Procedure 1. Show all endpoints to verify the endpoints to be changed:
Resources tab 255 SCSI Target
# scsitarget endpoint show list 2. Disable all endpoints: # scsitarget endpoint disable all 3. Delete the new, unnecessary endpoint, ep-new: # scsitarget endpoint del ep-new 4. Modify the endpoint you want to use, ep-1, by assigning it the new system address 10a: # scsitarget endpoint modify ep-1 system-address 10a 5. Enable all endpoints: # scsitarget endpoint enable all
Enabling an endpoint Enabling an endpoint enables the port only if it is currently disabled, that is, you are in non-NPIV mode. Procedure 1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Endpoints > Enable. If all endpoints are already enabled, a message to that effect is displayed. 3. In the Enable Endpoints dialog, select one or more endpoints from the list, and select Next. 4. After the confirmation, select Next to complete the task.
Disabling an endpoint Disabling an endpoint does not disable the associated port, unless all endpoints using the port are disabled, that is, you are in non- NPIV mode. Procedure 1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Endpoints > Disable. 3. In the Disable Endpoints dialog, select one or more endpoints from the list, and select Next. If an endpoint is in use, you are warned that disabling it might disrupt the system. 4. Select Next to complete the task.
Deleting an endpoint You may want to delete an endpoint if the underlying hardware is no longer available. However, if the underlying hardware is still present, or becomes available, a new endpoint for the hardware is discovered automatically and configured based on default values. Procedure 1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Endpoints > Delete. 3. In the Delete Endpoints dialog, select one or more endpoints from the list, and select Next. If an endpoint is in use, you are warned that deleting it might disrupt the system.
256 EMC Data Domain Operating System 5.7 Administration Guide SCSI Target
4. Select Next to complete the task.
Adding an initiator Add initiators to provide backup clients to connect to the system to read and write data using the FC (Fibre Channel) protocol. A specific initiator can support DD Boost over FC, or VTL, but not both. A maximum of 1024 initiators can be configured for a DD system. Procedure
1. Select Hardware > Fibre Channel > Resources. 2. Under Initiators, select Add (+ sign) 3. In the Add Initiator dialog, enter the port’s unique WWPN in the specified format. 4. Enter a Name for the initiator. 5. Select the Address Method: Auto is used for standard addressing, and VSA (Volume Set Addressing) is used primarily for addressing virtual buses, targets, and LUNs. 6. Select OK.
# scsitarget group add My_Group initiator My_Initiator
Modifying or deleting an initiator Before you can delete an initiator, it must be offline and not attached to any group. Otherwise, you will get an error message, and the initiator will not be deleted. You must delete all initiators in an access group before you can delete the access group. If an initiator remains visible, it may be automatically rediscovered. Procedure
1. Select Hardware > Fibre Channel > Resources. 2. Under Initiators, select one of the initiators. If you want to delete it, select Delete (X). If you want to modify it, select Modify (pencil) to display the Modify Initiator dialog. 3. Change the initiator’s Name and/or Address Method [Auto is used for standard addressing, and VSA (Volume Set Addressing) is used primarily for addressing virtual buses, targets, and LUNs.] 4. Select OK.
Recommendation to Set Initiator Aliases - CLI only It is strongly recommended that Initiator aliases be set to reduce confusion and human error during the configuration process. # vtl initiator set alias NewAliasName wwpn 21:00:00:e0:8b:9d:0b:e8 # vtl initiator show Initiator Group Status WWNN WWPN Port ------NewVTL aussie1 Online 20:00:00:e0:8b:9d:0b:e8 21:00:00:e0:8b:9d:0b:e8 6a Offline 20:00:00:e0:8b:9d:0b:e8 21:00:00:e0:8b:9d:0b:e8 6b
Initiator Symbolic Port Name Address Method ------NewVTL auto ------
Resources tab 257 SCSI Target
Setting a hard address (loop ID) Some backup software requires that all private-loop targets have a hard address (loop ID) that does not conflict with another node. The range for a loop ID is from 0 to 125. Procedure
1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Set Loop ID. 3. In the Set Loop ID dialog, enter the loop ID (from 0 to 125), and select OK.
Setting failover options You can set options for automatic failover and failback when NPIV is enabled. Here is the expected behavior for Fibre Channel port failover, by application:
l DD Boost-over-Fibre Channel operation is expected to continue without user intervention when the Fibre Channel endpoints failover. l VTL Fibre Channel operation is expected to be interrupted when the VTL Fibre Channel endpoints failover. You may need to perform discovery (that is, operating system discovery and configuration of VTL devices) on the initiators using the affected Fibre Channel endpoint. You should expect to re-start active backup and restore operations. l vDisk Fibre Channel operation is expected to continue without user intervention when the Fibre Channel endpoints failover. Automatic failback is not guaranteed if all ports are disabled and then subsequently enabled (which could be triggered by the administrator), as the order in which ports get enabled is unspecified. Procedure
1. Select Hardware > Fibre Channel > Resources. 2. Select More Tasks > Set Failover Options. 3. In the Set Failover Options dialog, enter the Failover and Failback Delay (in seconds) and whether to enable Automatic Failback, and select OK.
Access Groups tab The Hardware > Fibre Channel > Access Groups tab provides information about DD Boost and VTL access groups. Selecting the link to View DD Boost Groups or View VTL Groups takes you to the DD Boost or VTL pages.
Table 109 Access Groups
Group Name Name of access group. Service Service for this access group: either DD Boost or VTL. Endpoints Endpoints associated with this access group. Initiators Initiators associated with this access group. Number of Devices Number of devices associated with this access group.
258 EMC Data Domain Operating System 5.7 Administration Guide SCSI Target
Differences in FC link monitoring among DD OS versions Different releases of DD OS handle FC (Fibre Channel) Link Monitoring in different ways. DD OS 5.3 and later Port monitoring detects an FC port at system startup and raises an alert if the port is enabled and offline. To clear the alert, disable an unused port using the scsitarget port commands. DD OS 5.1 up to 5.3 If a port is offline, an alert notifies you that the link is down. This alert is managed, which means it stays active until cleared. This occurs when the VTL FC port is online or disabled. If the port is not in use, disable it unless it needs to be monitored. DD OS 5.0 up to 5.1 If a port is offline, an alert notifies you that the link is down. The alert is not managed, which means it does not stay active and does not appear in the current alerts list. When the port is online, an alert notifies you that the link is up. If the port is not in use, disable it unless it needs to be monitored. DD OS 4.9 up to 5.0 An FC port must be included in a VTL group to be monitored.
Differences in FC link monitoring among DD OS versions 259 SCSI Target
260 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 12 Working with DD Boost
This chapter includes:
l About Data Domain Boost software. 262 l Managing DD Boost with DD System Manager. 263 l About interface groups. 276 l Destroying DD Boost. 283 l Configuring DD Boost-over-Fibre Channel. 284 l Using DD Boost on HA systems. 288 l About the DD Boost tabs. 288
Working with DD Boost 261 Working with DD Boost
About Data Domain Boost software Data Domain Boost (DD Boost) software provides advanced integration with backup and enterprise applications for increased performance and ease of use. DD Boost distributes parts of the deduplication process to the backup server or application clients, enabling client-side deduplication for faster, more efficient backup and recovery. DD Boost software is an optional product that requires a separate license to operate on the Data Domain system. You can purchase a DD Boost software license key for a Data Domain system directly from EMC Data Domain.
A special license, BLOCK-SERVICES-PROTECTPOINT, is available to enable clients using ProtectPoint block services to have DD Boost functionality without a DD Boost license. If DD Boost is enabled for ProtectPoint clients only—that is, if only the BLOCK-SERVICES- PROTECTPOINT license is installed—the license status indicates that DD Boost is enabled for ProtectPoint only.
There are two components to DD Boost: one component that runs on the backup server and another that runs on the Data Domain system.
l In the context of the EMC NetWorker backup application, EMC Avamar backup application and other DDBoost partner backup applications, the component that runs on the backup server (DD Boost libraries) is integrated into the particular backup application. l In the context of Symantec backup applications (NetBackup and Backup Exec) and the Oracle RMAN plug-in, you need to download an appropriate version of the DD Boost plugin that is installed on each media server. The DD Boost plugin includes the DD Boost libraries for integrating with the DD Boost server running on the Data Domain system. The backup application (for example, Avamar, NetWorker, NetBackup, or Backup Exec) sets policies that control when backups and duplications occur. Administrators manage backup, duplication, and restores from a single console and can use all of the features of DD Boost, including WAN-efficient replicator software. The application manages all files (collections of data) in the catalog, even those created by the Data Domain system. In the Data Domain system, storage units that you create are exposed to backup applications that use the DD Boost protocol. For Symantec applications, storage units are viewed as disk pools. For Networker, storage units are viewed as logical storage units (LSUs). A storage unit is an MTree; therefore, it supports MTree quota settings. (Do not create an MTree in place of a storage unit.) This chapter does not contain installation instructions; refer to the documentation for the product you want to install. For example, for information about setting up DD Boost with Symantec backup applications (NetBackup and Backup Exec), see the EMC Data Domain Boost for OpenStorage Administration Guide. For information on setting up DD Boost with any other application, see the application-specific documentation. Additional information about configuring and managing DD Boost on the Data Domain system can also be found in the EMC Data Domain Boost for OpenStorage Administration Guide (for NetBackup and Backup Exec) and the EMC Data Domain Boost for Partner Integration Administration Guide (for other backup applications).
262 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
Managing DD Boost with DD System Manager Access the DD Boost view in System Manager. Procedure 1. Select Data Management > File System. Verify that the file system is enabled and running by checking its state. 2. Select Protocols > DD Boost. If you go to the DD Boost page without a license, the Status states that DD Boost is not licensed. Click Add License and enter a valid license in the Add License Key dialog box.
A special license, BLOCK-SERVICES-PROTECTPOINT, is available to enable clients using ProtectPoint block services to have DD Boost functionality without a DD Boost license. If DD Boost is enabled for ProtectPoint clients only—that is, if only the BLOCK- SERVICES-PROTECTPOINT license is installed—the license status indicates that DD Boost is enabled for ProtectPoint only.
Use the DD Boost tabs—Settings, Active Connections, IP Network, Fibre Channel, and Storage Units—to manage DD Boost.
Specifying DD Boost user names A DD Boost user is also a DD OS user. Specify a DD Boost user either by selecting an existing DD OS user name or by creating a new DD OS user name and making that name a DD Boost user. Backup applications use the DD Boost user name and password to connect to the Data Domain system. You must configure these credentials on each backup server that connects to this system. The Data Domain system supports multiple DD Boost users. For complete information about setting up DD Boost with Symantec NetBackup and Backup Exec, see the EMC Data Domain Boost for OpenStorage Administration Guide. For information on setting up DD Boost with other applications, see the EMC Data Domain Boost for Partner Integration Administration Guide and the application-specific documentation. Procedure 1. Select Protocols > DD Boost. 2. Select Add (+) above the Users with DD Boost Access list. The Add User dialog appears.
3. To select an existing user, select the user name in the drop-down list. EMC recommends that you select a user name with management role privileges set to none.
4. To create and select a new user, select Create a new Local User and do the following: a. Enter the new user name in the User field. The user must be configured in the backup application to connect to the Data Domain system.
Managing DD Boost with DD System Manager 263 Working with DD Boost
b. Enter the password twice in the appropriate fields. 5. Click Add.
Changing DD Boost user passwords Change a DD Boost user password. Procedure
1. Select Protocols > DD Boost > Settings. 2. Select a user in the Users with DD Boost Access list. 3. Click the Edit button (pencil icon) above the DD Boost user list. The Change Password dialog appears.
4. Enter the password twice in the appropriate boxes. 5. Click Change.
Removing a DD Boost user name Remove a user from the DD Boost access list. Procedure
1. Select Protocols > DD Boost > Settings. 2. Select the user in the Users with DD Boost Access list that needs to be removed. 3. Click Remove (X) above the DD Boost user list. The Remove User dialog appears.
4. Click Remove. After removal, the user remains in the DD OS access list.
Enabling DD Boost Use the DD Boost Settings tab to enable DD Boost and to select or add a DD Boost user. Procedure
1. Select Protocols > DD Boost > Settings. 2. Click Enable in the DD Boost Status area. The Enable DD Boost dialog box is displayed.
3. Select an existing user name from the menu, or add a new user by supplying the name, password, and role.
Disabling DD Boost Disabling DD Boost drops all active connections to the backup server. When you disable or destroy DD Boost, the DD Boost FC service is also disabled. Before you begin Ensure there are no jobs running from your backup application before disabling.
264 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
File replication started by DD Boost between two Data Domain restores is not canceled.
1. Select Protocols > DD Boost > Settings. 2. Click Disable in the DD Boost Status area. 3. Click OK in the Disable DD Boost confirmation dialog box.
Viewing DD Boost storage units Access the Storage Units tab to view and manage DD Boost storage units. The DD Boost Storage Unit tab:
l Lists the storage units and provides the following information for each storage unit:
Table 110 Storage unit information
Storage Unit The name of the storage unit. User The DD Boost user owning the storage unit. Quota Hard Limit The percentage of hard limit quota used. Last 24 hr Pre-Comp The amount of raw data from the backup application that has been written in the last 24 hours. Last 24 hr Post-Comp The amount of storage used after compression in the last 24 hours. Last 24 hr Comp Ratio The compression ratio for the last 24 hours. Weekly Avg Post-Comp The average amount of compressed storage used in the last five weeks. Last Week Post-Comp The average amount of compressed storage used in the last seven days. Weekly Avg Comp Ratio The average compression ratio for the last five weeks. Last Week Comp Ratio The average compression ratio for the last seven days.
l Allows you to create, modify, and delete storage units. l Displays four related tabs for a storage unit selected from the list: Storage Unit, Space Usage, Daily Written, and Data Movement.
The Data Movement tab is available only if the optional EMC Data Domain Extended Retention (formerly DD Archiver) license is installed.
l Takes you to Replication > On-Demand > File Replication when you click the View DD Boost Replications link.
Viewing DD Boost storage units 265 Working with DD Boost
A DD Replicator license is required for DD Boost to display tabs other than the File Replication tab.
Creating a storage unit You must create at least one storage unit on the Data Domain system, and a DD Boost user must be assigned to that storage unit. Use the Storage Units tab to create a storage unit. Each storage unit is a top-level subdirectory of the /data/col1 directory; there is no hierarchy among storage units. Procedure
1. Select Protocols > DD Boost > Storage Units. 2. Click Create (+). The Create Storage Unit dialog box is displayed.
3. Enter the storage unit name in the Name box. Each storage unit name must be unique. Storage unit names can be up to 50 characters. The following characters are acceptable:
l upper- and lower-case alphabetical characters: A-Z, a-z l numbers: 0-9 l embedded space
The storage-unit name must be enclosed in double quotes (") if the name has an embedded space.
l comma (,) l period (.), as long as it does not precede the name l exclamation mark (!) l number sign (#) l dollar sign ($) l per cent sign (%) l plus sign (+) l at sign (@) l equal sign (=) l ampersand (&) l semi-colon (;) l parenthesis [(and)] l square brackets ([and]) l curly brackets () l caret (^) l tilde (~)
266 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
l apostrophe (unslanted single quotation mark) l single slanted quotation mark (') l minus sign (-) l underscore (_)
4. To select an existing username that will have access to this storage unit, select the user name in the dropdown list. EMC recommends that you select a username with management role privileges set to none.
5. To create and select a new username that will have access to this storage unit, select Create a new Local User and: a. Enter the new user name in the User box. The user must be configured in the backup application to connect to the Data Domain system.
b. Enter the password twice in the appropriate boxes. 6. To set storage space restrictions to prevent a storage unit from consuming excess space: enter either a soft or hard limit quota setting, or both a hard and soft limit. With a soft limit an alert is sent when the storage unit size exceeds the limit, but data can still be written to it. Data cannot be written to the storage unit when the hard limit is reached.
Quota limits are pre-compressed values. To set quota limits, select Set to Specific Value and enter the value. Select the unit of measurement: MiB, GiB, TiB, or PiB.
When setting both soft and hard limits, a quota’s soft limit cannot exceed the quota’s hard limit.
7. Click Create. 8. Repeat the above steps for each Data Domain Boost-enabled system.
Viewing storage unit information From the DD Boost Storage Units tab, you can select a storage unit and access the Storage Unit, Space Usage, Daily Written, and Data Movement tabs for the selected storage unit. Storage Unit tab The Storage Unit tab shows detailed information for a selected storage unit in its Summary and Quota panels. The Snapshot panel shows snapshot details, allows you to create new snapshots and schedules, and provides a link to the Data Management > Snapshots tab.
l The Summary panel shows summarized information for the selected storage unit.
Viewing storage unit information 267 Working with DD Boost
Table 111 Summary panel
Summary item Description
Total Files The total number of file images on the storage unit. For compression details that you can download to a log file, click the Download Compression Details link. The generation can take up to several minutes. After it has completed, click Download. Full Path /data/col1/filename Tenant Unit Lists the tenant unit if one is assigned to the selected storage unit. Status R: read; W: write; Q: quota defined Pre-Comp Used The amount of pre-compressed storage already used.
l The Quota panel shows quota information for the selected storage unit. Table 112 Quota panel
Quota item Description
Quota Enforcement Enabled or disable. Clicking Quota takes you to the Data Management > Quota tab where you can configure quotas. Pre-Comp Soft Limit Current value of soft quota set for the storage unit. Pre-Comp Hard Limit Current value of hard quota set for the storage unit. Quota Summary Percentage of Hard Limit used.
To modify the pre-comp soft and hard limits shown in the tab: 1. Click the Configure button in the Quota panel. 2. In the Configure Quota dialog box, enter values for hard and soft quotas and select the unit of measurement: MiB, GiB, TiB, or PiB. Click OK.
l Snapshots The Snapshots panel shows information about the storage unit’s snapshots.
Table 113 Snapshots panel
Total Snapshots The total number of snapshots created for this MTree. A total of 750 snapshots can be created for each MTree. Expired The number of snapshots in this MTree that have been marked for deletion, but have not been removed with the clean operation as yet. Unexpired The number of snapshots in this MTree that are marked for keeping. Oldest Snapshot The date of the oldest snapshot for this MTree. Newest Snapshot The date of the newest snapshot for this MTree. Next Scheduled The date of the next scheduled snapshot.
268 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
Table 113 Snapshots panel (continued)
Assigned Snapshot The name of the snapshot schedule assigned to this MTree. Schedules
Using the Snapshots panel, you can:
n Assign a snapshot schedule to a selected storage unit: Click Assign Snapshot Schedules. Select the schedule’s checkbox; click OK and Close. n Create a new schedule: Click Assign Snapshot Schedules. Enter the new schedule’s name.
The snapshot name can be composed only of letters, numbers, _, -, %d (numeric day of the month: 01-31), %a (abbreviated weekday name), %m (numeric month of the year: 01-12), %b (abbreviated month name), %y (year, two digits), %Y (year, four digits), %H (hour: 00-23), and %M (minute: 00-59), following the pattern shown in the dialog box. Enter the new pattern and click Validate Pattern & Update Sample. Click Next. – Select when the schedule is to be executed: weekly, every day (or selected days), monthly on specific days that you select by clicking that date in the calendar, or on the last day of the month. Click Next. – Enter the times of the day when the schedule is to be executed: Either select At Specific Times or In Intervals. If you select a specific time, select the time from the list. Click Add (+) to add a time (24-hour format). For intervals, select In Intervals and set the start and end times and how often (Every), such as every eight hours. Click Next. – Enter the retention period for the snapshots in days, months, or years. Click Next. – Review the Summary of your configuration. Click Back to edit any of the values. Click Finish to create the schedule.
n Click the Snapshots link to go to the Data Management > Snapshots tab. Space Usage tab The Space Usage tab graph displays a visual representation of data usage for the storage unit over time. l Click a point on a graph line to display a box with data at that point. l Click Print (at the bottom on the graph) to open the standard Print dialog box. l Click Show in new window to display the graph in a new browser window. There are two types of graph data displayed: Logical Space Used (Pre-Compression) and Physical Capacity Used (Post-Compression). Daily Written tab The Daily Written view contains a graph that displays a visual representation of data that is written daily to the system over a period of time, selectable from 7 to 120 days. The data amounts are shown over time for pre- and post-compression amounts.
Viewing storage unit information 269 Working with DD Boost
Data Movement tab A graph in the same format as the Daily Written graph that shows the amount of disk space moved to the Extended Retention storage area (if the Extended Retention license is enabled).
Modifying a storage unit Use the Modify Storage Unit dialog to rename a storage unit, select a different existing user, create and select a new user, and edit quota settings. Procedure
1. Select Protocols > DD Boost > Storage Units. 2. In the Storage Unit list, select the storage unit to modify. 3. Click the pencil icon. The Modify Storage Unit dialog appears.
4. To rename the storage unit, edit the text in the Name field. 5. To select a different existing user, select the user name in the drop-down list. EMC recommends that you select a username with management role privileges set to none.
6. To create and select a new user, select Create a new Local User and do the following: a. Enter the new user name in the User box. The user must be configured in the backup application to connect to the Data Domain system.
b. Enter the password twice in the appropriate boxes. 7. Edit the Quota Settings as needed. To set storage space restrictions to prevent a storage unit from consuming excess space: enter either a soft or hard limit quota setting, or both a hard and soft limit. With a soft limit an alert is sent when the storage unit size exceeds the limit, but data can still be written to it. Data cannot be written to the storage unit when the hard limit is reached.
Quota limits are pre-compressed values. To set quota limits, select Set to Specific Value and enter the value. Select the unit of measurement: MiB, GiB, TiB, or PiB.
When setting both soft and hard limits, a quota’s soft limit cannot exceed the quota’s hard limit.
Renaming a storage unit Use the Modify Storage Unit dialog to rename a storage unit. Renaming a storage unit changes the name of the storage unit while retaining its:
270 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
l Username ownership l Stream limit configuration l Capacity quota configuration and physical reported size l AIR association on the local Data Domain system Procedure
1. Go to Protocols > DD Boost > Storage Units. 2. In the Storage Unit list, select the storage unit to rename. 3. Click the pencil icon. The Modify Storage Unit dialog appears.
4. Edit the text in the Name field. 5. Click Modify.
Deleting a storage unit Use the Storage Units tab to delete a storage unit from your Data Domain system. Deleting a storage unit removes the storage unit, as well as any images contained in the storage unit, from your Data Domain system. Procedure
1. Select Protocols > DD Boost > Storage Units. 2. Select the storage unit to be deleted from the list. 3. Click Delete (X). 4. Click OK. Results The storage unit is removed from your Data Domain system. You must also manually remove the corresponding backup application catalog entries.
Undeleting a storage unit Use the Storage Units tab to undelete a storage unit. Undeleting a storage unit recovers a previously deleted storage unit, including its:
l Username ownership l Stream limit configuration l Capacity quota configuration and physical reported size l AIR association on the local Data Domain system
Deleted storage units are available until the next filesys clean command is run.
1. Select Protocols > DD Boost > Storage Units > More Tasks > Undelete Storage Unit. 2. In the Undelete Storage Units dialog box, select the storage unit(s) that you want to undelete. 3. Click OK.
Deleting a storage unit 271 Working with DD Boost
Selecting DD Boost options Use the Set DD Boost Options dialog to specify settings for distributed segment processing, virtual synthetics, low bandwidth optimization for file replication, file replication encryption, and file replication network preference (IPv4 or IPv6). Procedure
1. To display the DD Boost option settings, select Protocols > DD Boost > Settings > Advanced Options. 2. To change the settings, select More Tasks > Set Options. The Set DD Boost Options dialog appears.
3. Select any option to be enabled. 4. Deselect any option to be disabled. To deselect a File Replication Network Preference option, select the other option.
You can also manage distributed segment processing via the ddboost option commands, which are described in detail in the EMC Data Domain Operating System Command Reference Guide.
Distributed segment processing Distributed segment processing increases backup throughput in almost all cases by eliminating duplicate data transmission between the media server and the Data Domain system. You can manage distributed segment processing via the ddboost option commands, which are described in detail in the EMC Data Domain Operating System Command Reference Guide.
Distributed segment processing is enabled by default with EMC Data Domain Extended Retention (formerly Data Domain Archiver) configurations and cannot be disabled.
Virtual synthetics A virtual synthetic full backup is the combination of the last full (synthetic or full) backup and all subsequent incremental backups. Virtual synthetics are enabled by default.
Low-bandwidth optimization If you use file replication over a low-bandwidth network (WAN), you can increase replication speed by using low bandwidth optimization. This feature provides additional compression during data transfer. Low bandwidth compression is available to Data Domain systems with an installed Replication license. Low-bandwidth optimization, which is disabled by default, is designed for use on networks with less than 6 Mbps aggregate bandwidth. Do not use this option if maximum file system write performance is required.
272 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
You can also manage low bandwidth optimization via the ddboost file- replication commands, which are described in detail in the EMC Data Domain Operating System Command Reference Guide.
File replication encryption You can encrypt the data replication stream by enabling the DD Boost file replication encryption option.
If DD Boost file replication encryption is used on systems without the Data at Rest option, it must be set to on for both the source and destination systems.
Managed file replication TCP port setting For DD Boost managed file replication, use the same global listen port on both the source and target Data Domain systems. To set the listen port, use the replication option command as described in the EMC Data Domain Operating Sysem Command Reference Guide.
File replication network preference Use this option to set the preferred network type for DD Boost file replication to either IPv4 or IPv6.
Managing certificates for DD Boost A host certificate allows DD Boost client programs to verify the identity of the system when establishing a connection. CA certificates identify certificate authorities that should be trusted by the system. The topics in this section describe how to manage host and CA certificates for DD Boost.
Adding a host certificate for DD Boost Add a host certificate to your system. DD OS supports one host certificate for DD Boost. Procedure 1. If you have not yet requested a host certificate, request one from a trusted CA. 2. When you have received a host certificate, copy or move it to the computer from which you run DD Service Manager. 3. Start DD System Manager on the system to which you want to add a host certificate.
DD System Manager supports certificate management only on the management system (which is the system running DD System Manager).
4. Select Protocols > DD Boost > More Tasks > Manage Certificates.
Managing certificates for DD Boost 273 Working with DD Boost
If you try to remotely manage certificates on a managed system, DD System Manager displays an information message at the top of the certificate management dialog. To manage certificates for a system, you must start DD System Manager on that system.
5. In the Host Certificate area, click Add. 6. To add a host certificate enclosed in a .p12 file, do the following: a. Select I want to upload the certificate as a .p12 file. b. Type the password in the Password box. c. Click Browse and select the host certificate file to upload to the system. d. Click Add. 7. To add a host certificate enclosed in a .pem file, do the following: a. Select I want to upload the public key as a .pem file and use a generated private key. b. Click Browse and select the host certificate file to upload to the system. c. Click Add.
Adding CA certificates for DD Boost Add a certificate for a trusted CA to your system. DD OS supports multiple certificates for trusted CAs. Procedure 1. Obtain a certificate for the trusted CA. 2. Copy or move the trusted CA certificate to the computer from which you run DD Service Manager. 3. Start DD System Manager on the system to which you want to add the CA certificate.
DD System Manager supports certificate management only on the management system (which is the system running DD System Manager).
4. Select Protocols > DD Boost > More Tasks > Manage Certificates.
If you try to remotely manage certificates on a managed system, DD System Manager displays an information message at the top of the certificate management dialog. To manage certificates for a system, you must start DD System Manager on that system.
5. In the CA Certificates area, click Add. The Add CA Certificate for DD Boost dialog appears.
6. To add a CA certificate enclosed in a .pem file, do the following: a. Select I want to upload the certificate as a .pem file. b. Click Browse, select the host certificate file to upload to the system, and click Open.
274 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
c. Click Add. 7. To add a CA certificate using copy and paste, do the following: a. Copy the certificate text to the clipboard using the controls in your operating system. b. Select I want to copy and paste the certificate text. c. Paste the certificate text in the box below the copy and paste selection. d. Click Add.
Managing DD Boost client access and encryption Use the DD Boost Settings tab to configure which specific clients, or set of clients, can establish a DD Boost connection with the Data Domain System and whether or not the client will use encryption. By default, the system is configured to allow all clients to have access, with no encryption.
Enabling in-flight encryption will impact system performance.
Adding a DD Boost client Create an allowed DD Boost client and specify whether the client will use encryption. Procedure 1. Select Protocols > DD Boost > Settings. 2. In the Allowed Clients section, click Create (+). The Add Allowed Client dialog appears.
3. Enter the hostname of the client. This can be a fully-qualified domain name (e.g. host1.emc.com) or a hostname with a wildcard (e.g. *.emc.com).
4. Select the Encryption Strength. The options are None (no encryption), Medium (AES128-SHA1), or High (AES256- SHA1).
5. Select the Authentication Mode. The options are One Way, Two Way, or Anonymous.
Modifying a DD Boost client Change the name, encryption strength, and authentication mode of an allowed DD Boost client. Procedure 1. Select Protocols > DD Boost > Settings. 2. In the Allowed Clients list, select the client to modify. 3. Click the Edit button, which displays a pencil icon. The Modify Allowed Client dialog appears.
Managing DD Boost client access and encryption 275 Working with DD Boost
4. To change the name of a client, edit the Client text. 5. To change the Encryption Strength, select the option. The options are None (no encryption), Medium (AES128-SHA1), or High (AES256- SHA1).
6. To change the Authentication Mode, select the option. The options are One Way, Two Way, or Anonymous.
Removing a DD Boost client Delete an allowed DD Boost client. Procedure 1. Select Protocols > DD Boost > Settings. 2. Select the client from the list. 3. Click Delete (X). The Delete Allowed Clients dialog appears.
4. Confirm and select the client name. Click OK.
About interface groups This feature lets you combine multiple Ethernet links into a group and register only one interface on the Data Domain system with the backup application. The DD Boost Library negotiates with the Data Domain system to obtain the best interface to send data. Load balancing provides higher physical throughput to the Data Domain system. Configuring an interface group creates a private network within the Data Domain system, comprised of the IP addresses designated as a group. Clients are assigned to a single group, and the group interface uses load balancing to improve data transfer performance and increase reliability. For example, in the Symantec NetBackup environment, media server clients use a single public network IP address to access the Data Domain system. All communication with the Data Domain system is initiated via this administered IP connection, which is configured on the NetBackup server. If an interface group is configured, when the Data Domain system receives data from the media server clients, the data transfer is load-balanced and distributed on all the interfaces in the group, providing higher input/output throughput, especially for customers who use multiple 1 GigE connections. The data transfer is load-balanced based on the number of connections outstanding on the interfaces. Only connections for backup and restore jobs are load-balanced. Check the Active Connections for more information on the number of outstanding connections on the interfaces in a group. Should an interface in the group fail, all the in-flight jobs to that interface are automatically resumed on healthy operational links (unbeknownst to the backup applications). Any jobs that are started subsequent to the failure are also routed to a healthy interface in the group. If the group is disabled or an attempt to recover on an alternate interface fails, the administered IP is used for recovery. Failure in one group will not utilize interfaces from another group. Consider the following information when managing interface groups.
276 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
l The IP address must be configured on the Data Domain system, and its interface enabled. To check the interface configuration, select Hardware > Ethernet > Interfaces page, and check for free ports. See the net chapter of the EMC Data Domain Operating System Command Reference Guide or the EMC Data Domain Operating System Initial Configuration Guide for information about configuring an IP address for an interface. l You can use the ifgroup commands to manage interface groups; these commands are described in detail in the EMC Data Domain Operating System Command Reference Guide. l Interface groups provide full support for static IPv6 addresses, providing the same capabilities for IPv6 as for IPv4. Concurrent IPv4 and IPv6 client connections are allowed. A client connected with IPv6 sees IPv6 ifgroup interfaces only. A client connected with IPv4 sees IPv4 ifgroup interfaces only. Individual ifgroups include all IPv4 addresses or all IPv6 addresses. For details, see the EMC Data Domain Boost Administration Guide. l Configured interfaces are listed in Active Connections, on the lower portion of the Activities page.
See Using DD Boost on HA systems on page 288 for important information about using interface groups with HA systems.
The topics that follow describe how to manage interface groups.
Interfaces IFGROUP supports physical and virtual interfaces. An IFGROUP interface is a member of a single IFGROUP and may consist of: l Physical interface such as eth0a l Virtual interface, created for link failover or link aggregation, such as veth1 l Virtual alias interface such as eth0a:2 or veth1:2 l Virtual VLAN interface such as eth0a.1 or veth1.1 l Within an IFGROUP , all interfaces must be on unique interfaces (Ethernet, virtual Ethernet) to ensure failover in the event of network error. IFGROUP provides full support for static IPv6 addresses, providing the same capabilities for IPv6 as for IPv4. Concurrent IPv4 and IPv6 client connections are allowed. A client connected with IPv6 sees IPv6 IFGROUP interfaces only. A client connected with IPv4 sees IPv4 IFGROUP interfaces only. Individual IFGROUPs include all IPv4 addresses or all IPv6 addresses. For more information, see the EMC Data Domain Boost Administration Guide.
Interface enforcement IFGROUP lets you enforce private network connectivity, ensuring that a failed job does not reconnect on the public network after network errors. When interface enforcement is enabled, a failed job can only retry on an alternative private network IP address. Interface enforcement is only available for clients that use IFGROUP interfaces. Interface enforcement is off (FALSE) by default. To enable interface enforcement, you must add the following setting to the system registry: system.ENFORCE_IFGROUP_RW=TRUE
Interfaces 277 Working with DD Boost
After you've made this entry in the registry, you must do a filesys restart for the setting to take effect. For more information, see the EMC Data Domain Boost Administration Guide.
Clients IFGROUP supports various naming formats for clients. Client selection is based on a specified order of precedence. An IFGROUP client is a member of a single ifgroup and may consist of:
l A fully qualified domain name (FQDN) such as ddboost.datadomain.com l Wild cards such as *.datadomain.com or “*”
l A short name for the client, such as ddboost l Client public IP range, such as 128.5.20.0/24 Prior to write or read processing, the client requests an IFGROUP IP address from the server. To select the client IFGROUP association, the client information is evaluated according to the following order of precedence. 1. IP address of the connected Data Domain system. If there is already an active connection between the client and the Data Domain system, and the connection exists on the interface in the IFGROUP, then the IFGROUP interfaces are made available for the client. 2. Connected client IP range. An IP mask check is done against the client source IP; if the client's source IP address matches the mask in the IFGROUP clients list, then the IFGROUP interfaces are made available for the client. l For IPv4, xx.xx.xx.0/24 provides a 24-bit mask against the connecting IP. The /24 represents what bits are masked when the client's source IP address is evaluated for access to the IFGROUP. For IPv4, 16, 20, 24, 28, and 32 bit masks are supported. l For IPv6, xxxx::0/112 provides a 112-bit mask against the connecting IP. The /112 represents what bits are masked when the client's source IP address is evaluated for access to the IFGROUP. For IPv6, 64, 112, and 128 bit masks are supported. This host-range check is useful for separate VLANs with many clients where there isn't a unique partial hostname (domain). 3. Client Name: abc-11.d1.com 4. Client Domain Name: *.d1.com 5. All Clients: * For more information, see the EMC Data Domain Boost for Partner Integration Administration Guide.
Creating interface groups Use the IP Network tab to create interface groups and to add interfaces and clients to the groups. Multiple interface groups improve the efficiency of DD Boost by allowing you to:
l Configure DD Boost to use specific interfaces configured into groups. l Assign clients to one of those interface groups.
278 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
l Monitor which interfaces are active with DD Boost clients. Create interface groups first, and then add clients (as new media servers become available) to an interface group. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Interface Groups section, click Add (+). 3. Enter the interface group name. 4. Select one or more interfaces. A maximum of 32 interfaces can be configured.
Depending upon aliasing configurations, some interfaces may not be selectable if they are sharing a physical interface with another interface in the same group. This is because each interface within the group must be on a different physical interface to ensure fail-over recovery.
5. Click OK. 6. In the Configured Clients section, click Add (+). 7. Enter a fully qualified client name or *.mydomain.com.
The * client is initially available to the default group. The * client may only be a member of one ifgroup.
8. Select a previously configured interface group, and click OK.
Enabling and disabling interface groups Use the IP Network tab to enable and disable interface groups. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Interface Groups section, select the interface group in the list.
If the interface group does not have both clients and interfaces assigned, you cannot enable the group.
3. Click Edit (pencil). 4. Click Enabled to enable the interface group; clear the checkbox to disable. 5. Click OK.
Modifying an interface group’s name and interfaces Use the IP Network tab to change an interface group's name and the interfaces associated with the group. Procedure
1. Select Protocols > DD Boost > IP Network.
Enabling and disabling interface groups 279 Working with DD Boost
2. In the Interface Groups section, select the interface group in the list. 3. Click Edit (pencil). 4. Retype the name to modify the name. The group name must be one to 24 characters long and contain only letters, numbers, underscores, and dashes. It cannot be the same as any other group name and cannot be “default”, “yes”, “no”, or “all.”
5. Select or deselect client interfaces in the Interfaces list.
If you remove all interfaces from the group, it will be automatically disabled.
Deleting an interface group Use the IP Network tab to delete an interface group. Deleting an interface group deletes all interfaces and clients associated with the group. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Interface Groups section, select the interface group in the list. The default group cannot be deleted. 3. Click Delete (X). 4. Confirm the deletion.
Adding a client to an interface group Use the IP Network tab to add clients to interface groups. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Configured Clients section, click Add (+). 3. Enter a name for the client. Client names must be unique and may consist of:
l * (for the default group only)
l Client public IP range: n For IPv4, xx.xx.xx.0/24 provides a 24-bit mask against the connecting IP. The /24 represents what bits are masked when the client's source IP address is evaluated for access to the IFGROUP. n For IPv6, xxxx::0/112 provides a 112-bit mask against the connecting IP. The /112 represents what bits are masked when the client's source IP address is evaluated for access to the IFGROUP. Client names have a maximum length of 128 characters. 4. Select a previously configured interface group, and click OK.
280 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
Modifying a client’s name or interface group Use the IP Network tab to change a client's name or interface group. Procedure 1. Select Protocols > DD Boost > IP Network. 2. In the Configured Clients section, select the client. 3. Click Edit (pencil). 4. Type a new client name. Client names must be unique and may consist of:
l FQDN l *.domain l * (for the default group only) l Client public IP range: n For IPv4, xx.xx.xx.0/24 provides a 24-bit mask against the connecting IP. The /24 represents what bits are masked when the client's source IP address is evaluated for access to the IFGROUP. n For IPv6, xxxx::0/112 provides a 112-bit mask against the connecting IP. The /112 represents what bits are masked when the client's source IP address is evaluated for access to the IFGROUP. Client names have a maximum length of 128 characters. 5. Select a new interface group from the menu.
The old interface group is disabled if it has no clients.
Deleting a client from the interface group Use the IP Network tab to delete a client from an interface group. Procedure 1. Select Protocols > DD Boost > IP Network. 2. In the Configured Clients section, select the client. 3. Click Delete (X).
If the interface group to which the client belongs has no other clients, the interface group is disabled.
4. Confirm the deletion.
Using interface groups for Managed File Replication (MFR) Interface groups can be used to control the interfaces used for DD Boost MFR, to direct the replication connection over a specific network, and to use multiple network interfaces
Modifying a client’s name or interface group 281 Working with DD Boost
with high bandwidth and reliability for failover conditions. All Data Domain IP types are supported—IPv4 or IPv6, Alias IP/VLAN IP, and LACP/failover aggregation.
Interface groups used for replication are different from the interface groups previously explained and are supported for DD Boost Managed File Replication (MFR) only. For detailed information about using interface groups for MFR, see the EMC Data Domain Boost for Partner Integration Administration Guide or the EMC Data Domain Boost for OpenStorage Administration Guide.
Without the use of interface groups, configuration for replication requires several steps: 1. Adding an entry in the /etc/hosts file on the source Data Domain system for the target Data Domain system and hard coding one of the private LAN network interfaces as the destination IP address. 2. Adding a route on the source Data Domain system to the target Data Domain system specifying a physical or virtual port on the source Data Domain system to the remote destination IP address. 3. Configuring LACP through the network on all switches between the Data Domain systems for load balancing and failover. 4. Requiring different applications to use different names for the target Data Domain system to avoid naming conflicts in the /etc/hosts file. Using interface groups for replication simplifies this configuration through the use of the DD OS System Manager or DD OS CLI commands. Using interface groups to configure the replication path lets you:
l Redirect a hostname-resolved IP address away from the public network, using another private Data Domain system IP address. l Identify an interface group based on configured selection criteria, providing a single interface group where all the interfaces are reachable from the target Data Domain system. l Select a private network interface from a list of interfaces belonging to a group, ensuring that the interface is healthy. l Provide load balancing across multiple Data Domain interfaces within the same private network. l Provide a failover interface for recovery for the interfaces of the interface group. l Provide host failover if configured on the source Data Domain system. l Use Network Address Translation (NAT) The selection order for determining an interface group match for file replication is: 1. Local MTree (storage-unit) path and a specific remote Data Domain hostname 2. Local MTree (storage-unit) path with any remote Data Domain hostname 3. Any MTree (storage-unit) path with a specific Data Domain hostname The same MTree can appear in multiple interface groups only if it has a different Data Domain hostname. The same Data Domain hostname can appear in multiple interface groups only if it has a different MTree path. The remote hostname is expected to be an FQDN, such as dd890-1.emc.com. The interface group selection is performed locally on both the source Data Domain system and the target Data Domain system, independent of each other. For a WAN
282 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
replication network, only the remote interface group needs to be configured since the source IP address corresponds to the gateway for the remote IP address.
Adding a replication path to an interface group Use the IP Network tab to add replication paths to interface groups. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Configured Replication Paths section, click Add (+). 3. Enter values for MTree and/or Remote Host. 4. Select a previously configured interface group, and click OK.
Modifying a replication path for an interface group Use the IP Network tab to modify replication paths for interface groups. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Configured Replication Paths section, select the replication path. 3. Click Edit (pencil). 4. Modify any or all values for MTree, Remote Host, or Interface Group. 5. Click OK.
Deleting a replication path for an interface group Use the IP Network tab to delete replication paths for interface groups. Procedure
1. Select Protocols > DD Boost > IP Network. 2. In the Configured Replication Paths section, select the replication path. 3. Click Delete (X). 4. In the Delete Replication Path(s) dialog, click OK.
Destroying DD Boost Use this option to permanently remove all of the data (images) contained in the storage units. When you disable or destroy DD Boost, the DD Boost FC service is also disabled. Only an administrative user can destroy DD Boost. Procedure 1. Manually remove (expire) the corresponding backup application catalog entries.
If multiple backup applications are using the same Data Domain system, then remove all entries from each of those applications’ catalogs.
2. Select Protocols > DD Boost > More Tasks > Destroy DD Boost. 3. Enter your administrative credentials when prompted. 4. Click OK.
Destroying DD Boost 283 Working with DD Boost
Configuring DD Boost-over-Fibre Channel In earlier versions of DD OS, all communication between the DD Boost Library and any Data Domain system was performed using IP networking. DD OS now offers Fibre Channel as an alternative transport mechanism for communication between the DD Boost Library and the Data Domain system.
Windows, Linux, HP-UX (64-bit Itanium architecture), AIX, and Solaris client environments are supported.
Enabling DD Boost users Before you can configure the DD Boost-over-FC service on a Data Domain system, you must add one or more DD Boost users and enable DD Boost. Before you begin
l Log in to System Manager. For instructions, see “Logging In and Out of DD System Manager.” CLI equivalent login as: sysadmin Data Domain OS 5.7.x.x-12345 Using keyboard-interactive authentication. Password: l If you are using the CLI, ensure that the SCSI target daemon is enabled: # scsitarget enable Please wait . SCSI Target subsystem is enabled.
If you are using System Manager, the SCSI target daemon is automatically enabled when you enable the DD Boost-over-FC service (later in this procedure).
l Verify that the DD Boost license is installed. In System Manager, select Protocols > DD Boost > Settings. If the Status indicates that DD Boost is not licensed, click Add License and enter a valid license in the Add License Key dialog box. CLI equivalents # license show # license add license-code Procedure
1. Select Protocols > DD Boost > Settings. 2. In the Users with DD Boost Access section, specify one or more DD Boost user names. A DD Boost user is also a DD OS user. When specifying a DD Boost user name, you can select an existing DD OS user name, or you can create a new DD OS user name and make that name a DD Boost user. This release supports multiple DD Boost users. For detailed instructions, see “Specifying DD Boost User Names.”
# user add username [password password]
284 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
# ddboost set user-name exampleuser
3. Click Enable to enable DD Boost. CLI equivalent
# ddboost enable Starting DDBOOST, please wait. DDBOOST is enabled.
Results You are now ready to configure the DD Boost-over-FC service on the Data Domain system.
Configuring DD Boost After you have added user(s) and enabled DD Boost, you need to enable the Fibre Channel option and specify the DD Boost Fibre Channel server name. Depending on your application, you may also need to create one or more storage units and install the DD Boost API/plug-in on media servers that will access the Data Domain system. Procedure
1. Select Protocols > DD Boost > Fibre Channel. 2. Click Enable to enable Fibre Channel transport. CLI equivalent
# ddboost option set fc enabled Please wait. DD Boost option "FC" set to enabled.
3. To change the DD Boost Fibre Channel server name from the default (hostname), click Edit, enter a new server name, and click OK. CLI equivalent
# ddboost fc dfc-server-name set DFC-ddbeta2 DDBoost dfc-server-name is set to "DFC-ddbeta2" for DDBoost FC. Configure clients to use "DFC-DFC-ddbeta2" for DDBoost FC.
4. Select Protocols > DD Boost > Storage Units to create a storage unit (if not already created by the application). You must create at least one storage unit on the Data Domain system, and a DD Boost user must be assigned to that storage unit. For detailed instructions, see “Creating a Storage Unit.” CLI equivalent
# ddboost storage-unit create storage_unit_name-su
5. Install the DD Boost API/plug-in (if necessary, based on the application). The DD Boost OpenStorage plug-in software must be installed on NetBackup media servers that need to access the Data Domain system. This plug-in includes the required DD Boost Library that integrates with the Data Domain system. For detailed installation and configuration instructions, see the Data Domain Boost for OpenStorage Administration Guide or the Data Domain Boost Administration Guide.
Configuring DD Boost 285 Working with DD Boost
Results You are now ready to verify connectivity and create access groups.
Verifying connectivity and creating access groups Go to Hardware > Fibre Channel > Resources to manage initiators and endpoints for access points. Go to Protocols > DD Boost > Fibre Channel to create and manage DD Boost-over-FC access groups.
Avoid making access group changes on a Data Domain system during active backup or restore jobs. A change may cause an active job to fail. The impact of changes during active jobs depends on a combination of backup software and host configurations.
1. Select Hardware > Fibre Channel > Resources > Initiators to verify that initiators are present. It is recommended that you assign aliases to initiators to reduce confusion during the configuration process. CLI equivalent # scsitarget initiator show list Initiator System Address Group Service ------initiator-1 21:00:00:24:ff:31:b7:16 n/a n/a initiator-2 21:00:00:24:ff:31:b8:32 n/a n/a initiator-3 25:00:00:21:88:00:73:ee n/a n/a initiator-4 50:06:01:6d:3c:e0:68:14 n/a n/a initiator-5 50:06:01:6a:46:e0:55:9a n/a n/a initiator-6 21:00:00:24:ff:31:b7:17 n/a n/a initiator-7 21:00:00:24:ff:31:b8:33 n/a n/a initiator-8 25:10:00:21:88:00:73:ee n/a n/a initiator-9 50:06:01:6c:3c:e0:68:14 n/a n/a initiator-10 50:06:01:6b:46:e0:55:9a n/a n/a tsm6_p23 21:00:00:24:ff:31:ce:f8 SetUp_Test VTL ------
2. To assign an alias to an initiator, select one of the initiators and click the pencil (edit) icon. In the Name field of the Modify Initiator dialog, enter the alias and click OK. CLI equivalents # scsitarget initiator rename initiator-1 initiator-renamed Initiator 'initiator-1' successfully renamed. # scsitarget initiator show list Initiator System Address Group Service ------initiator-2 21:00:00:24:ff:31:b8:32 n/a n/a initiator-renamed 21:00:00:24:ff:31:b7:16 n/a n/a ------
3. On the Resources tab, verify that endpoints are present and enabled. CLI equivalent # scsitarget endpoint show list ------endpoint-fc-0 5a FibreChannel Yes Online endpoint-fc-1 5b FibreChannel Yes Online ------
286 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
4. Go to Protocols > DD Boost > Fibre Channel. 5. In the DD Boost Access Groups area, click the + icon to add an access group. 6. Enter a unique name for the access group. Duplicate names are not supported. CLI equivalent # ddboost fc group create test-dfc-group DDBoost FC Group "test-dfc-group" successfully created.
7. Select one or more initiators. Optionally, replace the initiator name by entering a new one. Click Next. CLI equivalent #ddboost fc group add test-dfc-group initiator initiator-5 Initiator(s) "initiator-5" added to group "test-dfc-group".
An initiator is a port on an HBA attached to a backup client that connects to the system for the purpose of reading and writing data using the Fibre Channel protocol. The WWPN is the unique World-Wide Port Name of the Fibre Channel port in the media server. 8. Specify the number of DD Boost devices to be used by the group. This number determines which devices the initiator can discover and, therefore, the number of I/O paths to the Data Domain system. The default is one, the minimum is one, and the maximum is 64. CLI equivalent # ddboost fc group modify Test device-set count 5 Added 3 devices.
See the Data Domain Boost for OpenStorage Administration Guide for the recommended value for different clients. 9. Indicate which endpoints to include in the group: all, none, or select from the list of endpoints. Click Next. CLI equivalents # scsitarget group add Test device ddboost-dev8 primary-endpoint all secondary-endpoint all Device 'ddboost-dev8' successfully added to group. # scsitarget group add Test device ddboost-dev8 primary-endpoint endpoint-fc-1 secondary-endpoint fc-port-0 Device 'ddboost-dev8' is already in group 'Test'.
When presenting LUNs via attached FC ports on HBAs, ports can be designated as primary, secondary or none. A primary port for a set of LUNs is the port that is currently advertizing those LUNs to a fabric. A secondary port is a port that will broadcast a set of LUNs in the event of primary path failure (this requires manual intervention). A setting of none is used in the case where you do not wish to advertize selected LUNs. The presentation of LUNs is dependent upon the SAN topology. 10. Review the Summary and make any modifications. Click Finish to create the access group, which is displayed in the DD Boost Access Groups list. CLI equivalent # scsitarget group show detailed
Verifying connectivity and creating access groups 287 Working with DD Boost
To change settings for an existing access group, select it from the list and click the pencil icon (Modify).
Deleting access groups Use the Fibre Channel tab to delete access groups. Procedure
1. Select Protocols > DD Boost > Fibre Channel. 2. Select the group to be deleted from the DD Boost Access Groups list.
You cannot delete a group that has initiators assigned to it. Edit the group to remove the initiators first.
3. Click Delete (X).
Using DD Boost on HA systems HA provides seamless failover of any application using DD Boost—that is, any backup or restore operation continues with no manual intervention required. All other DD Boost user scenarios are supported on HA systems as well, including managed file replication (MFR), distributed segment processing (DSP), filecopy, and dynamic interface groups (DIG). Note these special considerations for using DD Boost on HA systems:
l On HA-enabled Data Domain systems, failovers of the DD server occur in less than 10 minutes. However, recovery of DD Boost applications may take longer than this, because Boost application recovery cannot begin until the DD server failover is complete. In addition, Boost application recovery cannot start until the application invokes the Boost library. l DD Boost on HA systems requires that the Boost applications be using Boost HA libraries; applications using non-HA Boost libraries do not see seamless failover. l MFR will fail over seamlessly when both the source and destination systems are HA- enabled. MFR is also supported on partial HA configurations (that is, when either the source or destination system is enabled, but not both) when the failure occurs on the HA-enabled system. For more information, see the EMC DD Boost for OpenStorage Administration Guide or the EMC DD Boost for Partner Integration Administration Guide. l Dynamic interface groups should not include IP addresses associated with the direct interconnection between the active and standby Data Domain systems. l DD Boost clients must be configured to use floating IP addresses.
About the DD Boost tabs Learn to use the DD Boost tabs in System Manager.
288 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
Settings Use the Settings tab to enable or disable DD Boost, select clients and users, and specify advanced options. The Settings tab shows the DD Boost status (Enabled or Disabled). Use the Status button to switch between Enabled or Disabled. Under Allowed Clients, select the clients that are to have access to the system. Use the Add, Modify, and Delete buttons to manage the list of clients. Under Users with DD Boost Access, select the users that are to have DD Boost access. Use the Add, Change Password, and Remove buttons to manage the list of users. Expand Advanced Options to see which advanced options are enabled. Go to More Tasks > Set Options to reset these options.
Active Connections Use the Active Connections tab to see information about clients, interfaces, and outbound files.
Table 114 Connected client information
Client The name of the connected client. Idle Whether the client is idle (Yes) or not (No). CPUs The number of CPUs that the client has, such as 8. Memory (GiB) The amount of memory (in GiB) the client has, such as 7.8. Plug-In Version The DD Boost plug-in version installed, such as 2.2.1.1. OS Version The operating system version installed, such as Linux 2.6.1 7-1.2142_FC4smp x86_64. Application Version The backup application version installed, such as NetBackup 6.5.6. Encrypted Whether the connection is encrypted (Yes) or not (No). DSP Whether or not the connection is using Distributed Segment Processing (DSP) or not. Transport Type of transport being used, such as IPv4, IPv6 or DFC (Fibre Channel).
Table 115 Configured interface connection information
Interface The IP address of the interface. Interface Group One of the following:
l The name of the interface group.
l None, if not a member of one.
Backup The number of active backup connections.
Settings 289 Working with DD Boost
Table 115 Configured interface connection information (continued)
Restore The number of active restore connections. Replication The number of active replication connections. Synthetic The number of synthetic backups. Total The total number of connections for the interface.
Table 116 Outbound file replication information
Outbound files item Description
File Name The name of the outgoing image file.
Target Host The name of the host receiving the file. Logical Bytes to Transfer The number of logical bytes to be transferred. Logical Bytes Transferred The number of logical bytes already transferred. Low Bandwidth Optimization The number of low-bandwidth bytes already transferred.
IP Network The IP Network tab lists configured interface groups. Details include whether or not a group is enabled and any configured client interfaces. Administrators can use the Interface Group menu to view which clients are associated with an interface group.
Fibre Channel The Fibre Channel tab lists configured DD Boost access groups. Use the Fibre Channel tab to create and delete access groups and to configure initiators, devices, and endpoints for DD Boost access groups.
Storage Units Use the Storage Unit tab to create, modify, and delete storage units. To see detailed information about a listed storage unit, select its name.
Table 117 Storage unit: Detailed information
Existing Storage Units Storage Unit Name The name of the storage unit. Pre-Comp Used The amount of pre-compressed storage already used. Pre-Comp Soft Limit Current value of soft quota set for the storage unit. % of Pre-Comp Soft Limit Used Percentage of hard limit quota used. Pre-Comp Hard Limit Current value of hard quota set for the storage unit.
290 EMC Data Domain Operating System 5.7 Administration Guide Working with DD Boost
Table 117 Storage unit: Detailed information (continued)
% of Pre-Comp Hard Limit Used Percentage of hard limit quota used. Storage Unit Details Select the storage unit in the list. Total Files The total number of file images on the storage unit. Download Files Link to download storage unit file details in .tsv format. You must allow pop-ups to use this function. Compression Ratio The compression ratio achieved on the files. Metadata Size The amount of space used for metadata information. Storage Unit Status The current status of the storage unit (combinations are supported). Status can be:
l RD—Replication destination
l RLE—Retention lock enabled
l RLD—Retention lock disabled
Quota Enforcement Click Quota to go to the Data Management Quota page, which lists hard and soft quota values/percentage used by MTrees. Quota Summary Percentage of Hard Limit used. Original Size The size of the file before compression was performed. Global Compression Size The total size after global compression of the files in the storage unit when they were written. Locally Compressed Size Total size after local compression of the files in the storage unit when they were written.
Storage Units 291 Working with DD Boost
292 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 13 DD Virtual Tape Library
This chapter includes:
l DD Virtual Tape Library overview. 294 l Planning a VTL. 294 l Managing a VTL. 299 l Working with libraries. 303 l Working with a selected library. 306 l Viewing changer information. 314 l Working with drives. 315 l Working with a selected drive. 317 l Working with tapes. 317 l Working with the vault. 319 l Working with access groups. 319 l Working with a selected access group. 323 l Working with resources. 325 l Working with pools. 330 l Working with a selected pool. 332
DD Virtual Tape Library 293 DD Virtual Tape Library
DD Virtual Tape Library overview EMC Data Domain Virtual Tape Library (VTL) is a disk-based backup system that emulates the use of physical tapes. It enables backup applications to connect to and manage DD system storage using functionality almost identical to a physical tape library. Virtual tape drives are accessible to backup software in the same way as physical tape drives. After you create these drives in a VTL, they appear to the backup software as SCSI tape drives. The VTL, itself, appears to the backup software as a SCSI robotic device accessed through standard driver interfaces. However, the backup software (not the DD system that is configured as a VTL) manages the movement of the media changer and backup images. The following terms have special meaning when used with VTL: l Library: A library emulates a physical tape library with drives, changer, CAPs (cartridge access ports), and slots (cartridge slots). l Tape: A tape is represented as a file. Tapes can be imported from the vault to a library. Tapes can be exported from a library to the vault. Tapes can be moved within a library across drives, slots, and CAPs. l Pool: A pool is a collection of tapes that maps to a directory on the file system. Pools are used to replicate tapes to a destination. You can convert directory-based pools to MTree-based pools to take advantage of the greater functionality of MTrees. l Vault: The vault holds tapes not being used by any library. Tapes reside in either a library or the vault. VTL has been tested with, and is supported by, specific backup software and hardware configurations. For more information, see the appropriate Backup Compatibility Guide on the EMC Online Support Site. VTL supports simultaneous use of the tape library and file system (NFS/CIFS/DD Boost) interfaces. When DR (disaster recovery) is needed, pools and tapes can be replicated to a remote DD system using the DD Replicator. To protect data on tapes from modification, tapes can be locked using Retention Lock Governance software.
At present, for 16 Gb/s, EMC supports fabric and point-to-point topologies. Other topologies will present issues.
Planning a VTL The VTL (Virtual Tape Library) feature has very specific requirements, such as proper licensing, interface cards, user permissions, etc. These requirements are listed here, complete with details and recommendations. l An appropriate VTL license. n VTL is a licensed feature, and you must use NDMP (Network Data Management Protocol) over IP (Internet Protocol) or VTL directly over FC (Fibre Channel). n An additional license is required for IBM i systems – the I/OS license. n Adding a VTL license through the DD System Manager automatically disables and enables the VTL feature.
294 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
l An installed FC interface card or VTL configured to use NDMP.
n If the VTL communication between a backup server and a DD system is through an FC interface, the DD system must have an FC interface card installed. Notice that whenever an FC interface card is removed from (or changed within) a DD system, any VTL configuration associated with that card must be updated. n If the VTL communication between a backup server and a DD system is through NDMP, no FC interface card is required. However, you must configure the TapeServer access group. Also, when using NDMP, all initiator and port functionality does not apply. l A backup software minimum record (block) size.
n EMC strongly recommends that backup software be set to use a minimum record (block) size of 64 KiB or larger. Larger sizes usually give faster performance and better data compression. n Depending on your backup application, if you change the size after the initial configuration, data written with the original size might become unreadable. l Appropriate user access to the system.
n For basic tape operations and monitoring, only a user login is required. n To enable and configure VTL services and perform other configuration tasks, a sysadmin login is required.
VTL limits Before setting up or using a VTL, review these limits on size, slots, etc.
l I/O Size – The maximum supported I/O size for any DD system using VTL is 1 MB. l Libraries – VTL supports a maximum of 64 libraries per DD system (that is, 64 VTL instances on each DD system). l Initiators – VTL supports a maximum of 1024 initiators or WWPNs (world-wide port names) per DD system. l Tape Drives – Information about tape drives is presented in the next section. l Data Streams – Information about data streams is presented in the following table. Table 118 Data streams sent to a Data Domain system
Model RAM/ NVRAM Backup Backup Repla Repla dest Mixed write read source streams streams streams streams
DD140, DD160, 4 GB or 6 GB / 16 4 15 20 w VTL. Status In the Virtual Tape Libraries > VTL Service area, you can see the status of your VTL process is displayed at the top, for example, Enabled: Running. The first part of the status will be Enabled (on) or Disabled (off). The second part will be one of the following process states.
Table 122 VTL process states
Running VTL process is enabled and active (shown in green). Starting VTL process is starting. Stopping VTL process is being shut down. Stopped VTL process is disabled (shown in red). Timing out VTL process crashed and is attempting an automatic restart. Stuck After several failed automatic restarts, the VTL process is unable to shut down normally, so an attempt is being made to kill it.
VTL License The VTL License line tells you whether your VTL license has been applied. If it says Unlicensed, select Add License. Enter your license key in the Add License Key dialog. Select Next and OK.
All license information should have been populated as part of the factory configuration process; however, if the VTL option was purchased later, the VTL license key may not have been available at that time.
CLI Equivalent You can also verify that the VTL license has been installed at the CLI: # license show ## License Key Feature ------1 DEFA-EFCD-FCDE-CDEF Replication 2 EFCD-FCDE-CDEF-DEFA VTL ------If the license is not present, each unit comes with documentation – a quick install card – which will show the licenses that have been purchased. Enter the following command to populate the license key. # license add license-code
I/OS License (for IBM i users) For customers of IBM i, the I/OS License line tells you whether your I/OS license has been applied. If it says Unlicensed, select Add License. You must enter a valid I/OS license in either of these formats: XXXX-XXXX-XXXX-XXXX or XXXX-XXXX-XXXX-XXXX-XXXX. Your
300 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
I/OS license must be installed before creating a library and drives to be used on an IBM i system. Select Next and OK.
Enabling VTL Enabling VTL broadcasts the WWN of the Data Domain HBA to customer fabric and enables all libraries and library drives. If a forwarding plan is required in the form of change control processes, this process should be enabled to facilitate zoning. Procedure 1. Make sure that you have a VTL license and that the file system is enabled. 2. Select Virtual Tape Libraries > VTL Service. 3. To the right of the Status area, select Enable. 4. In the Enable Service dialog, select OK. 5. After VTL has been enabled, note that Status will change to Enabled: Running in green. Also note that the configured VTL options are displayed in the Option Defaults area.
CLI Equivalent # vtl enable Starting VTL, please wait . VTL is enabled.
Disabling VTL Disabling VTL closes all libraries and shuts down the VTL process. Procedure
1. Select Virtual Tape Libraries > VTL Service. 2. To the right of the Status area, select Disable. 3. In the Disable Service dialog, select OK. 4. After VTL has been disabled, notice that the Status has changed to Disabled: Stopped in red.
CLI Equivalent # vtl disable
VTL option defaults The Option Default area of the VTL Service page displays the current settings for default VTL options (auto-eject, auto-offline, and barcode-length) that you can configure. In the Virtual Tape Libraries > VTL Service area, the current default options for your VTL are displayed. Select Configure to change any of these values.
Table 123 Option Defaults
Property Lists the configured options:
Enabling VTL 301 DD Virtual Tape Library
Table 123 Option Defaults (continued)
Value Provides the value for each configured option:
l auto-eject: default (disabled), enabled, or disabled
l auto-offline: default (disabled), enabled, or disabled
l barcode-length: default (8), 6, or 8
Configuring VTL default options You can configure VTL default options when you add a license, create a library, or any time thereafter.
VTLs are assigned global options, by default, and those options are updated whenever global options change, unless you change them manually using this method.
1. Select Virtual Tape Libraries > VTL Service. 2. In the Option Defaults area, select Configure. In the Configure Default Options dialog, change any or all of the default options.
Table 124 VTL default options
Option Values Notes
auto-eject default (disabled), Enabling auto-eject causes any tape put into a CAP (cartridge enable, or disable access port) to automatically move to the virtual vault, unless:
l the tape came from the vault, in which case the tape stays in the CAP. l an ALLOW_MEDIUM_REMOVAL command with a 0 value (false) has been issued to the library to prevent the removal of the medium from the CAP to the outside world.
auto- default (disabled), Enabling auto-offline takes a drive offline automatically offline enable, or disable before a tape move operation is performed.
barcode- default (8), 6 or 8 Although a VTL barcode consists of 8 characters, either 6 or length [automatically set to 8 characters may be transmitted to a backup application, 6 for L180, depending on the changer type. RESTORER-L180, and DDVTL changer models]
302 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
3. Select OK. 4. Or to disable all of these service options, select Reset to Factory, and the values will be immediately reset to factory defaults.
Working with libraries A library emulates a physical tape library with drives, changer, CAPs (cartridge access ports), and slots (cartridge slots). Selecting Virtual Tape Libraries > VTL Service > Libraries displays detailed information for all configured libraries.
Table 125 Library information
Name The name of a configured library. Drives The number of drives configured in the library.
Slots The number of slots configured in the library. CAPs The number of CAPs (cartridge access ports) configured in the library.
From the More Tasks menu, you can create and delete libraries, as well as search for tapes.
Creating libraries VTL supports a maximum of 64 libraries per system, that is, 64 concurrently active virtual tape library instances on each DD system. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries. 2. Select More Tasks > Library > Create 3. In the Create Library dialog, enter the following information:
Table 126 Create Library dialog
Field User input
Library Name Enter a name of from 1 to 32 alphanumeric characters. Number of Drives Enter the number of drives (from 1 to 98 (see Note). The number of drives to be created will correspond to the number of data streams that will write to a library.
The maximum number of drives supported by a VTL depends on the number of CPU cores and the amount of memory installed (both RAM and NVRAM, if applicable) on a DD system.
Drive Model Select the desired model from the drop-down list:
Working with libraries 303 DD Virtual Tape Library
Table 126 Create Library dialog (continued)
Field User input
l IBM-LTO-5 (default)
l HP-LTO-4 Do not mix drive types, or media types, in the same library. This can cause unexpected results and/or errors in the backup operation. Number of Slots Enter the number of slots in the library. Here are some things to consider:
l The number of slots must be equal to or greater than the number of drives.
l You can have up to 32,000 slots per individual library
l You can have up to 64,000 slots per system.
l Try to have enough slots so tapes remain in the VTL and never need to be exported to a vault – to avoid reconfiguring the VTL and to ease management overhead.
l Consider any applications that are licensed by the number of slots. As an example, for a standard 100-GB cartridge on a DD580, you might configure 5000 slots. This would be enough to hold up tp 500 TB (assuming reasonably compressible data).
Number of CAPs (Optional) Enter the number of cartridge access ports (CAPs).
l You can have up to 100 CAPs per library.
l You can have up to 1000 CAPs per system. Check your particular backup software application documentation on the EMC Online Support Site for guidance. Changer Model Name Select the desired model from the drop-down list:
l TS3500 (which should be used for IBMi deployments)
l DDVTL Check your particular backup software application documentation on the EMC Online Support Site for guidance. Also refer to the VTL support matrix to see the compatibility of emulated libraries to supported software. Options
304 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Table 126 Create Library dialog (continued)
Field User input
auto-eject default (disabled), enable, disable auto-offline default (disabled), enable, disable barcode-length default (8), 6, 8 [automatically set to 6 for L180, RESTORER-L180, and DDVTL changer models]
4. Select OK. After the Create Library status dialog shows Completed, select OK. The new library appears under the Libraries icon in the VTL Service tree, and the options you have configured appear as icons under the library. Selecting the library displays details about the library in the Information Panel. Note that access to VTLs and drives is managed with Access Groups.
CLI Equivalent # vtl add NewVTL model L180 slots 50 caps 5 This adds the VTL library, NewVTL. Use 'vtl show config NewVTL' to view it.
# vtl drive add NewVTL count 4 model IBM-LTO-3 This adds 4 IBM-LTO-3 drives to the VTL library, NewVTL.
Deleting libraries When a tape is in a drive within a library, and that library is deleted, the tape is moved to the vault. However, the tape's pool does not change. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries. 2. Select More Tasks > Library > Delete. 3. In the Delete Libraries dialog, select or confirm the checkbox of the items to delete:
l The name of each library, or l Library Names, to delete all libraries
4. Select Next. 5. Verify the libraries to delete, and select Submit in the confirmation dialogs. 6. After the Delete Libraries Status dialog shows Completed, select Close. The selected libraries are deleted from the VTL.
CLI Equivalent # vtl del OldVTL
Deleting libraries 305 DD Virtual Tape Library
Searching for tapes You can use a variety of criteria – location, pool, and/or barcode – to search for a tape. Procedure 1. Select Virtual Tape Libraries or Pools. 2. Select the area to search (library, vault, pool). 3. Select More Tasks > Tapes > Search. 4. In the Search Tapes dialog, enter information about the tape(s) you want to find.
Table 127 Search Tapes dialog
Field User input
Location Specify a location, or leave the default (All).
Pool Select the name of the pool in which to search for the tape. If no pools have been created, use the Default pool. Barcode Specify a unique barcode. or leave the default (*) to return a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters. Count Enter the maximum number of tapes you want to be returned to you. If you leave this blank, the barcode default (*) is used.
5. Select Search.
Working with a selected library Selecting Virtual Tape Libraries > VTL Service > Libraries > library displays detailed information for a selected library.
Table 128 Devices
Device The elements in the library, such a drives, slots, and CAPs (cartridge access ports). Loaded The number of devices with media loaded. Empty The number of devices with no media loaded. Total The total number of loaded and empty devices.
Table 129 Options
auto-eject enabled or disabled auto-offline enabled or disabled barcode-length 6 or 8
306 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Table 129 Options (continued)
Table 130 Tapes
Pool The name of the pool where the tapes are located. Tape Count The number of tapes in that pool. Capacity The total configured data capacity of the tapes in that pool, in GiB (Gibibytes, the base-2 equivalent of GB, Gigabytes). Used The amount of space used on the virtual tapes in that pool. Average Compression The average amount of compression achieved on the data on the tapes in that pool.
From the More Tasks menu, you can delete, rename, or set options for a library; create, delete, import, export, or move tapes; and add or delete slots and CAPs.
Creating tapes You can create tapes in either a library or a pool. If initiated from a pool, the system first creates the tapes, then imports them to the library. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries > library or Vault or Pools > Pools > pool. 2. Select More Tasks > Tapes > Create. 3. In the Create Tapes dialog, enter the following information about the tape:
Table 131 Create Tapes dialog
Field User input
Library (if initiated If a drop-down menu is enabled, select the library or leave the default from a library) selection. Pool Name Select the name of the pool in which the tape will reside, from the drop- down list. If no pools have been created, use the Default pool. Number of Tapes For a library, select from 1 to 20. For a pool, select from 1 to 100,000, or leave the default (20). [Although the number of supported tapes is unlimited, you can create no more than 100,000 tapes at a time.] Starting Barcode Enter the initial barcode number (using the format A99000LA). Tape Capacity (optional) Specify the number of GiBs from 1 to 4000 for each tape (this setting overrides the barcode capacity setting). For efficient use of disk space, use 100 GiB or fewer.
4. Select OK and Close.
Creating tapes 307 DD Virtual Tape Library
# vtl tape add A00000L1 capacity 100 count 5 pool VTL_Pool . added 5 tape(s).
You must auto-increment tape volume names in base10 format.
Deleting tapes You can delete tapes from either a library or a pool. If initiated from a library, the system first exports the tapes, then deletes them. The tapes must be in the vault, not in a library. On a Replication destination DD system, deleting a tape is not permitted. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries > library or Vault or Pools > Pools > pool. 2. Select More Tasks > Tapes > Delete. 3. In the Delete Tapes dialog, enter search information about the tapes to delete, and select Search:
Table 132 Delete Tapes dialog
Field User input
Location If there is a drop-down list, select a library, or leave the default Vault selection. Pool Select the name of the pool in which to search for the tape. If no pools have been created, use the Default pool. Barcode Specify a unique barcode, or leave the default (*) to search for a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters. Count Enter the maximum number of tapes you want to be returned to you. If you leave this blank, the barcode default (*) is used. Tapes Per Select the maximum number of tapes to display per page – possible values are 15, Page 30, and 45.
Select all Select the Select All Pages checkbox to select all tapes returned by the search pages query. Items Shows the number of tapes selected across multiple pages – updated Selected automatically for each tape selection.
4. Select the checkbox of the tape that should be deleted or the checkbox on the heading column to delete all tapes, and select Next. 5. Select Submit in the confirmation window, and select Close.
After a tape is removed, the physical disk space used for the tape is not reclaimed until after a file system cleaning operation.
308 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
# vtl tape del barcode [count count] [pool pool] For example:
# vtl tape del A00000L1
You can act on ranges; however, if there is a missing tape in the range, the action will stop.
Importing tapes Importing a tape means that an existing tape will be moved from the vault to a library slot, drive, or cartridge access port (CAP). The number of tapes you can import at one time is limited by the number of empty slots in the library, that is, you cannot import more tapes than the number of currently empty slots. To view the available slots for a library, select the library from the stack menu. The information panel for the library shows the count in the Empty column.
l If a tape is in a drive, and the tape origin is known to be a slot, a slot is reserved. l If a tape is in a drive, and the tape origin is unknown (slot or CAP), a slot is reserved. l If a tape is in a drive, and the tape origin is known to be a CAP, a slot is not reserved. (The tape returns to the CAP when removed from the drive.) l To move a tape to a drive, see the section on moving tapes, which follows. Procedure 1. You can import tapes using either step a. or step b.
a. Select Virtual Tape Libraries > VTL Service > Libraries > library. Then, select More Tasks > Tapes > Import. In the Import Tapes dialog, enter search information about the tapes to import, and select Search:
Table 133 Import Tapes dialog
Field User input
Location If there is a drop-down list, select the location of the tape, or leave the default of Vault. Pool Select the name of the pool in which to search for the tape. If no pools have been created, use the Default pool. Barcode Specify a unique barcode. or leave the default (*) to return a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters. Count Enter the maximum number of tapes you want to be returned to you. If you leave this blank, the barcode default (*) is used. Tapes Per Select the maximum number of tapes to display per page. Possible values Page are 15, 30, and 45. Items Shows the number of tapes selected across multiple pages – updated Selected automatically for each tape selection.
Importing tapes 309 DD Virtual Tape Library
Table 133 Import Tapes dialog (continued)
Based on the previous conditions, a default set of tapes is searched to select the tapes to import. If pool, barcode, or count is changed, select Search to update the set of tapes available from which to choose.
b. Select Virtual Tape Libraries > VTL Service > Libraries> library > Changer > Drives > drive > Tapes. Select tapes to import by selecting the checkbox next to:
l An individual tape, or l The Barcode column to select all tapes on the current page, or l The Select all pages checkbox to select all tapes returned by the search query. Only tapes showing Vault in the Location can be imported. Select Import from Vault. This button is disabled by default and enabled only if all of the selected tapes are from the Vault.
2. From the Import Tapes: library view, verify the summary information and the tape list, and select OK. 3. Select Close in the status window.
# vtl tape show pool VTL_Pool Processing tapes. Barcode Pool Location State Size Used (%) Comp ModTime ------A00000L3 VTL_Pool vault RW 100 GiB 0.0 GiB (0.00%) 0x 2010/07/16 09:50:41 A00001L3 VTL_Pool vault RW 100 GiB 0.0 GiB (0.00%) 0x 2010/07/16 09:50:41 A00002L3 VTL_Pool vault RW 100 GiB 0.0 GiB (0.00%) 0x 2010/07/16 09:50:41 A00003L3 VTL_Pool vault RW 100 GiB 0.0 GiB (0.00%) 0x 2010/07/16 09:50:41 A00004L3 VTL_Pool vault RW 100 GiB 0.0 GiB (0.00%) 0x 2010/07/16 09:50:41 ------VTL Tape Summary ------Total number of tapes: 5 Total pools: 1 Total size of tapes: 500 GiB Total space used by tapes: 0.0 GiB Average Compression: 0.0x
# vtl import NewVTL barcode A00000L3 count 5 pool VTL_Pool . imported 5 tape(s).
# vtl tape show pool VTL_Pool Processing tapes.
VTL Tape Summary ------Total number of tapes: 5 Total pools: 1 Total size of tapes: 500 GiB Total space used by tapes: 0.0 GiB Average Compression: 0.0x
310 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Exporting tapes Exporting a tape removes that tape from a slot, drive, or cartridge-access port (CAP) and sends it to the vault. Procedure 1. You can export tapes using either step a. or step b. a. Select Virtual Tape Libraries > VTL Service > Libraries > library. Then, select More Tasks > Tapes > Export. In the Export Tapes dialog, enter search information about the tapes to export, and select Search:
Table 134 Export Tapes dialog
Field User input
Location If there is a drop-down list, select the name of the library where the tape is located, or leave the selected library. Pool Select the name of the pool in which to search for the tape. If no pools have been created, use the Default pool. Barcode Specify a unique barcode. or leave the default (*) to return a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters. Count Enter the maximum number of tapes you want to be returned to you. If you leave this blank, the barcode default (*) is used. Tapes Per Select the maximum number of tapes to display per page. Possible values are Page 15, 30, and 45.
Select all Select the Select All Pages checkbox to select all tapes returned by the pages search query. Items Shows the number of tapes selected across multiple pages – updated Selected automatically for each tape selection.
b. Select Virtual Tape Libraries > VTL Service > Libraries> library > Changer > Drives > drive > Tapes. Select tapes to export by selecting the checkbox next to:
l An individual tape, or l The Barcode column to select all tapes on the current page, or l The Select all pages checkbox to select all tapes returned by the search query. Only tapes with a library name in the Location column can be exported. Select Export from Library. This button is disabled by default and enabled only if all of the selected tapes have a library name in the Location column.
2. From the Export Tapes: library view, verify the summary information and the tape list, and select OK. 3. Select Close in the status window.
Moving tapes between devices within a library Tapes can be moved between physical devices within a library to mimic backup software procedures for physical tape libraries (which move a tape in a library from a slot to a
Exporting tapes 311 DD Virtual Tape Library
drive, a slot to a CAP, a CAP to a drive, and the reverse). In a physical tape library, backup software never moves a tape outside the library. Therefore, the destination library cannot change and is shown only for clarification. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries library . Note that when started from a library, the Tapes panel allows tapes to be moved only between devices.
2. Select More Tasks > Tapes > Move. Note that when started from a library, the Tapes panel allows tapes to be moved only between devices.
3. In the Move Tape dialog, enter search information about the tapes to move, and select Search:
Table 135 Move Tape dialog
Field User input
Location Location cannot be changed. Pool N/A Barcode Specify a unique barcode. or leave the default (*) to return a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters. Count Enter the maximum number of tapes you want to be returned to you. If you leave this blank, the barcode default (*) is used. Tapes Per Select the maximum number of tapes to display per page. Possible values are 15, Page 30, and 45. Items Shows the number of tapes selected across multiple pages – updated Selected automatically for each tape selection.
4. From the search results list, select the tape or tapes to move. 5. Do one of the following: a. Select the device from the Device list (for example, a slot, drive, or CAP), and enter a starting address using sequential numbers for the second and subsequent tapes. For each tape to be moved, if the specified address is occupied, the next available address is used. b. Leave the address blank if the tape in a drive originally came from a slot and is to be returned to that slot; or if the tape is to be moved to the next available slot. 6. Select Next. 7. In the Move Tape dialog, verify the summary information and the tape listing, and select Submit. 8. Select Close in the status window.
312 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Adding slots You can add slots from a configured library to change the number of storage elements.
Some backup applications do not automatically recognize that slots have been added to a VTL. See your application documentation for information on how to configure the application to recognize this type of change.
1. Select Virtual Tape Libraries > VTL Service > Libraries > library. 2. Select More Tasks > Slots > Add. 3. In the Add Slots dialog, enter the Number of Slots to add. The total number of slots in a library, or in all libraries on a system, cannot exceed 32,000 for a library and 64,000 for a system. 4. Select OK and Close when the status shows Completed.
Deleting slots You can delete slots from a configured library to change the number of storage elements.
Some backup applications do not automatically recognize that slots have been deleted from a VTL. See your application documentation for information on how to configure the application to recognize this type of change.
Procedure 1. If the slot that you want to delete contains cartridges, move those cartridges to the vault. The system will delete only empty, uncommitted slots. 2. Select Virtual Tape Libraries > VTL Service > Libraries > library. 3. Select More Tasks > Slots > Delete. 4. In the Delete Slots dialog, enter the Number of Slots to delete. 5. Select OK and Close when the status shows Completed.
Adding CAPs You can add CAPs (cartridge access ports) from a configured library to change the number of storage elements.
CAPs are used by a limited number of backup applications. See your application documentation to ensure that CAPs are supported.
1. Select Virtual Tape Libraries > VTL Service > Libraries > library. 2. Select More Tasks > CAPs > Add. 3. In the Add CAPs dialog, enter the Number of CAPs to add. You can add from 1 to 100 CAPs per library and from 1 to 1,000 CAPs per system.
Adding slots 313 DD Virtual Tape Library
4. Select OK and Close when the status shows Completed.
Deleting CAPs You can delete CAPs (cartridge access ports) from a configured library to change the number of storage elements.
Some backup applications do not automatically recognize that CAPs have been deleted from a VTL. See your application documentation for information on how to configure the application to recognize this type of change.
Procedure 1. If the CAP that you want to delete contains cartridges, move those cartridges to the vault, or this will be done automatically. 2. Select Virtual Tape Libraries > VTL Service > Libraries > library. 3. Select More Tasks > CAPs > Delete. 4. In the Delete CAPs dialog, enter the Number of CAPs to delete. You can delete a maximum of 100 CAPs per library or 1000 CAPs per system. 5. Select OK and Close when the status shows Completed.
Viewing changer information There can be only one changer per VTL. The changer model you select depends on your specific configuration. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries . 2. Select a specific library. 3. If not expanded, select the plus sign (+) on the left to open the library, and select a Changer element to display the Changer information panel, which provides the following information.
Table 136 Changer information panel
Vendor The name of the vendor who manufactured the changer Product The model name Revision The revision level Serial Number The changer serial number
314 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Working with drives Selecting Virtual Tape Libraries > VTL Service > Libraries > library > Drives displays detailed information for all drives for a selected library.
Table 137 Drives information panel
Drive The list of drives by name, where name is “Drive #” and # is a number between 1 and n representing the address or location of the drive in the list of drives. Vendor The manufacturer or vendor of the drive, for example, IBM. Product The product name of the drive, for example, ULTRIUM-TD5. Revision The revision number of the drive product.
Serial The serial number of the drive product. Number Status Whether the drive is Empty, Open, Locked, or Loaded. A tape must be present for the drive to be locked or loaded. Tape The barcode of the tape in the drive (if any). Pool The pool of the tape in the drive (if any).
Tape and library drivers – To work with drives, you must use the tape and library drivers supplied by your backup software vendor that support the IBM LTO-1, IBM LTO-2, IBM LTO-3, IBM LTO-4, IBM LTO-5 (default), HP-LTO-3, or HP-LTO-4 drives and the StorageTek L180 (default), RESTORER-L180, IBM TS3500, I2000, I6000, or DDVTL libraries. For more information, see the Application Compatibility Matrices and Integration Guides for your vendors. When configuring drives, also keep in mind the limits on backup data streams, which are determined by the platform in use. LTO drive capacities – Because the DD system treats LTO drives as virtual drives, you can set a maximum capacity to 4 TiB (4000 GiB) for each drive type. The default capacities for each LTO drive type are as follows:
l LTO-1 drive: 100 GiB l LTO-2 drive: 200 GiB l LTO-3 drive: 400 GiB l LTO-4 drive: 800 GiB l LTO-5 drive: 1.5 TiB Migrating LTO-1 tapes – You can migrate tapes from existing LTO-1 type VTLs to VTLs that include other supported LTO-type tapes and drives. The migration options are different for each backup application, so follow the instructions in the LTO tape migration guide specific to your application. To find the appropriate guide, go to the EMC Online Support Site, and in the search text box, type in LTO Tape Migration for VTLs. Tape full: Early warning – You will receive a warning when the remaining tape space is almost completely full, that is, greater than 99.9, but less than 100 percent. The application can continue writing until the end of the tape to reach 100 percent capacity. The last write, however, is not recoverable. From the More Tasks menu, you can create or delete a drive.
Working with drives 315 DD Virtual Tape Library
Creating drives See the Number of drives supported by a VTL section to determine the maximum number of drives supported for your particular VTL. Procedure
1. Select Virtual Tape Libraries > VTL Service > Libraries > library> Changer > Drives. 2. Select More Tasks > Drives > Create. 3. In the Create Drive dialog, enter the following information:
Table 138 Create Drive dialog
Field User input
Location Select a library name, or leave the name selected.
Number of See the table in the Number of Drives Supported by a VTL section, earlier in this Drives chapter. Model Name Select the model from the drop-down list. If another drive already exists, this option is inactive, and the existing drive type must be used. You cannot mix drive types in the same library.
l IBM-LTO-5 (default)
4. Select OK, and when the status shows Completed, select OK. The added drive appears in the Drives list.
Deleting drives A drive must be empty before it can be deleted. Procedure 1. If there is a tape in the drive that you want to delete, remove the tape. 2. Select Virtual Tape Libraries > VTL Service > Libraries > library > Changer > Drives. 3. Select More Tasks > Drives > Delete. 4. In the Delete Drives dialog, select the checkboxes of the drives to delete, or select the Drive checkbox to delete all drives. 5. Select Next, and after verifying that the correct drive(s) has been selected for deletion, select Submit. 6. When the Delete Drive Status dialog shows Completed, select Close. The drive will have been removed from the Drives list.
316 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Working with a selected drive Selecting Virtual Tape Libraries > VTL Service > Libraries > library > Drives > drive displays detailed information for a selected drive.
Table 139 Drive Tab
Drive The list of drives by name, where name is “Drive #” and # is a number between 1 and n representing the address or location of the drive in the list of drives. Vendor The manufacturer or vendor of the drive, for example, IBM. Product The product name of the drive, for example, ULTRIUM-TD5.
Revision The revision number of the drive product. Serial Number The serial number of the drive product. Status Whether the drive is Empty, Open, Locked, or Loaded. A tape must be present for the drive to be locked or loaded. Tape The barcode of the tape in the drive (if any). Pool The pool of the tape in the drive (if any).
Table 140 Statistics Tab
Endpoint The specific name of the endpoint. Ops/s The operations per second. Read KiB/s The speed of reads in KiB per second. Write KiB/s The speed of writes in KiB per second.
From the More Tasks menu, you can delete the drive or perform a refresh.
Working with tapes A tape is represented as a file. Tapes can be imported from the vault to a library. Tapes can be exported from a library to the vault. Tapes can be moved within a library across drives, slots (cartridge slots), and CAPs (cartridge access ports). When tapes are created, they are placed into the vault. After they have been added to the vault, they can be imported, exported, moved, searched, or removed. Selecting Virtual Tape Libraries > VTL Service > Libraries> library >Tapes displays detailed information for all tapes for a selected library.
Working with a selected drive 317 DD Virtual Tape Library
Table 141 Tape description
Barcode The unique barcode for the tape. Pool The name of the pool that holds the tape. The Default pool holds all tapes unassigned to a user-created pool. Location The location of the tape - whether in a library (and which drive, CAP, or slot number) or in the virtual vault. State The state of the tape:
l RW – Read-writable
l RL – Retention-locked
l RO – Readable only
l WP – Write-protected
l RD – Replication destination
Capacity The total capacity of the tape. Used The amount of space used on the tape. Compression The amount of compression performed on the data on a tape. Last Modified The date of the last change to the tape’s information. Modification times used by the system for age-based policies might differ from the last modified time displayed in the tape information sections of the DD System Manager. Locked Until If a Retention Lock deadline has been set, the time set is shown. If no retention lock exists, this value is Not specified.
From the information panel, you can import a tape from the vault, export a tape to the library, set a tape's state, create a tape, or delete a tape. From the More Tasks menu, you can move a tape.
Changing a tape’s write or retention lock state Before changing a tape's write or retention lock state, the tape must have been created and imported. VTL tapes follow the standard Data Domain retention lock policy. After the retention period for a tape has expired, it cannot be written to or changed (however, it can be deleted). Procedure 1. Select Virtual Tape Libraries > VTL Service > Libraries > library > Tapes. 2. Select the tape to modify from the list, and select Set State (above the list). 3. In the Set Tape State dialog, select Read-Writeable, Write-Protected, or Retention- Lock. 4. If the state is Retention-Lock, either l enter the tape’s expiration date in a specified number of days, weeks, months, years, or l select the calendar icon, and select a date from the calendar. The Retention-Lock expires at noon on the selected date.
318 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
5. Select Next, and select Submit to change the state.
Working with the vault The vault holds tapes not being used by any library. Tapes reside in either a library or the vault. Selecting Virtual Tape Libraries > VTL Service > Vault displays detailed information for the Default pool and any other existing pools in the vault.
Table 142 Pool Information
Location The name of the pool. Type Whether it is a Directory or MTree pool.
Tape Count The number of tapes in the pool. Capacity The total amount of space in the pool. Used The amount of space used on in the pool. Average Compression The average amount of compression in the pool.
From the More Tasks menu, you can create, delete, and search for tapes in the vault.
Working with access groups Access groups hold a collection of initiator WWPNs (worldwide port names) or aliases and the drives and changers they are allowed to access. A VTL default group named TapeServer lets you add devices that will support NDMP (Network Data Management Protocol)-based backup applications. Access group configuration allows initiators (in general backup applications) to read and write data to devices in the same access group. Access groups let clients access only selected LUNs (media changers or virtual tape drives) on a system. A client set up for an access group can access only devices in its access group. Avoid making access group changes on a DD system during active backup or restore jobs. A change may cause an active job to fail. The impact of changes during active jobs depends on a combination of backup software and host configurations. Selecting Access Groups > Groups displays the following information for all access groups.
Table 143 Access group information
Group Name Name of group. Initiators Number of initiators in group. Devices Number of devices in group.
If you select View All Access Groups, you are taken to the Fibre Channel view.
Working with the vault 319 DD Virtual Tape Library
From the More Tasks menu, you can create or delete a group.
Creating an access group Access groups manage access between devices and initiators. Do not use the default TapeServer access group unless you are using NDMP. Procedure
1. Select Access Groups > Groups. 2. Select More Tasks > Group > Create 3. In the Create Access Group dialog, enter a name, from 1 to 128 characters, and select Next. 4. Add devices, and select Next. 5. Review the summary, and select Finish or Back, as appropriate. CLI Equivalent
# scsitarget group create My_Group service My_Service
Adding an access group device Access group configuration allows initiators (in general backup applications) to read and write data to devices in the same access group. Procedure
1. Select Access Groups > Groups. You can also select a specific group. 2. Select More Tasks > Group > Create or Group > Configure. 3. In the Create or Modify Access Group dialog, enter or modify the Group Name if desired. (This field is required.) 4. To configure initiators to the access group, check the box next to the initiator. You can add initiators to the group later. 5. Select Next. 6. In the Devices display, select Add (+) to display the Add Devices dialog. a. Verify that the correct library is selected in the Library Name drop-down list, or select another library. b. In the Device area, select the checkboxes of the devices (changer and drives) to be included in the group. c. Optionally, specify a starting LUN in the LUN Start Address text box. This is the LUN that the DD system returns to the initiator. Each device is uniquely identified by the library and the device name. (For example, it is possible to have drive 1 in Library 1 and drive 1 in Library 2). Therefore, a LUN is associated with a device, which is identified by its library and device name. When presenting LUNs via attached FC ports on FC HBA/SLIC, ports can be designated as primary, secondary, or none. A Primary port for a set of LUNs is the port that is currently advertizing those LUNs to a fabric. A secondary port is a port that will broadcast a set of LUNs in the event of primary path failure (this requires manual intervention). A setting of none is used in the case where you do not wish to advertize selected LUNs. The presentation of LUNs depends on the SAN topology in question.
320 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
The initiators in the access group interact with the LUN devices that are added to the group. The maximum LUN accepted when creating an access group is 16383. A LUN can be used only once for an individual group. The same LUN can be used with multiple groups. Some VTL initiators (clients) have specific rules for VTL target LUN numbering; for example, requiring LUN 0 or requiring contiguous LUNs. If these rules are not followed, an initiator may not be able to access some or all of the LUNs assigned to a VTL target port. Check your initiator documentation for special rules, and if necessary, alter the device LUNs on the VTL target port to follow the rules. For example, if an initiator requires LUN 0 to be assigned on the VTL target port, check the LUNs for devices assigned to ports, and if there is no device assigned to LUN 0, change the LUN of a device so it is assigned to LUN 0.
d. In the Primary and Secondary Endpoints area, select an option to determine from which ports the selected device will be seen. The following conditions apply for designated ports: l all – The checked device is seen from all ports. l none – The checked device is not seen from any port. l select – The checked device is to be seen from selected ports. Select the checkboxes of the appropriate ports. If only primary ports are selected, the checked device is visible only from primary ports. If only secondary ports are selected, the checked device is visible only from secondary ports. Secondary ports can be used if the primary ports become unavailable. The switchover to a secondary port is not an automatic operation. You must manually switch the VTL device to the secondary ports if the primary ports become unavailable. The port list is a list of physical port numbers. A port number denotes the PCI slot and a letter denotes the port on a PCI card. Examples are 1a, 1b, or 2a, 2b. A drive appears with the same LUN on all the ports that you have configured.
e. Select OK. You are returned to the Devices dialog box where the new group is listed. To add more devices, repeat these five substeps. 7. Select Next. 8. Select Close when the Completed status message is displayed.
# vtl group add VTL_Group vtl NewVTL changer lun 0 primary-port all secondary-port all # vtl group add VTL_Group vtl NewVTL drive 1 lun 1 primary-port all secondary-port all # vtl group add SetUp_Test vtl SetUp_Test drive 3 lun 3 primary-port endpoint-fc-0 secondary-port endpoint-fc-1
# vtl group show Setup_Test Group: SetUp_Test
Initiators: Initiator Alias Initiator WWPN
Creating an access group 321 DD Virtual Tape Library
Devices: Device Name LUN Primary Ports Secondary Ports In-use Ports ------SetUp_Test changer 0 all all all SetUp_Test drive 1 1 all all all SetUp_Test drive 2 2 5a 5b 5a SetUp_Test drive 3 3 endpoint-fc-0 endpoint-fc-1 endpoint-fc-0 ------
Modifying or deleting an access group device You may need to modify or delete a device from an access group. Procedure 1. Select Protocols > VTL > Access Groups > Groups > group. 2. Select More Tasks > Group > Configure. 3. In the Modify Access Group dialog, enter or modify the Group Name. (This field is required.) 4. To configure initiators to the access group, check the box next to the initiator. You can add initiators to the group later. 5. Select Next. 6. Select a device, and select the edit (pencil) icon to display the Modify Devices dialog. Then, follow steps a-e. If you simply want to delete the device, select the delete (X) icon, and skip to step e. a. Verify that the correct library is selected in the Library drop-down list, or select another library. b. In the Devices to Modify area, select the checkboxes of the devices (Changer and drives) to be modified. c. Optionally, modify the starting LUN (logical unit number) in the LUN Start Address box. This is the LUN that the DD system returns to the initiator. Each device is uniquely identified by the library and the device name. (For example, it is possible to have drive 1 in Library 1 and drive 1 in Library 2). Therefore, a LUN is associated with a device, which is identified by its library and device name. The initiators in the access group interact with the LUN devices that are added to the group. The maximum LUN accepted when creating an access group is 16383. A LUN can be used only once for an individual group. The same LUN can be used with multiple groups. Some VTL initiators (clients) have specific rules for VTL target LUN numbering; for example, requiring LUN 0 or requiring contiguous LUNs. If these rules are not followed, an initiator may not be able to access some or all of the LUNs assigned to a VTL target port. Check your initiator documentation for special rules, and if necessary, alter the device LUNs on the VTL target port to follow the rules. For example, if an initiator requires LUN 0 to be assigned on the VTL target port, check the LUNs for devices assigned to ports, and if there is no device assigned to LUN 0, change the LUN of a device so it is assigned to LUN 0.
322 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
d. In the Primary and Secondary Ports area, change the option that determines the ports from which the selected device is seen. The following conditions apply for designated ports:
l all – The checked device is seen from all ports. l none – The checked device is not seen from any port. l select – The checked device is seen from selected ports. Select the checkboxes of the ports from which it will be seen. If only primary ports are selected, the checked device is visible only from primary ports. If only secondary ports are selected, the checked device is visible only from secondary ports. Secondary ports can be used if primary ports become unavailable. The switchover to a secondary port is not an automatic operation. You must manually switch the VTL device to the secondary ports if the primary ports become unavailable. The port list is a list of physical port numbers. A port number denotes the PCI slot, and a letter denotes the port on a PCI card. Examples are 1a, 1b, or 2a, 2b. A drive appears with the same LUN on all ports that you have configured.
Deleting an access group Before you can delete an access group, you must remove all of its initiators and LUNs. Procedure 1. Remove all of the initiators and LUNs from the group. 2. Select Access Groups > Groups. 3. Select More Tasks > Group > Delete. 4. In the Delete Group dialog, select the checkbox of the group to be removed, and select Next. 5. In the groups confirmation dialog, verify the deletion, and select Submit. 6. Select Close when the Delete Groups Status displays Completed.
# scsitarget group destroy My_Group
Working with a selected access group Selecting Access Groups > Groups > group displays the following information for a selected access group.
Table 144 LUNs tab
LUN Device address – maximum number is 16383. A LUN can be used only once within a group, but can be used again within
Deleting an access group 323 DD Virtual Tape Library
Table 144 LUNs tab (continued)
another group. VTL devices added to a group must use contiguous LUNs. Library Name of library associated with LUN. Device Changers and drives. In-Use Endpoints Set of endpoints currently being used: primary or secondary. Primary Endpoints Initial (or default) endpoint used by backup application. In the event of a failure on this endpoint, the secondary endpoints may be used, if available. Secondary Endpoints Set of fail-over endpoints to use if primary endpoint fails.
Table 145 Initiators tab
Name Name of initiator, which is either the WWPN or the alias assigned to the initiator. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel port.
From the More Tasks menu, with a group selected, you can configure that group, or set endpoints in use.
Selecting endpoints for a device Since endpoints connect a device to an initiator, use this process to set up the endpoints before you connect the device. Procedure
1. Select Access Groups > Groups > group. 2. Select More Tasks > Endpoints > Set In-Use. 3. In the Set in-Use Endpoints dialog, select only specific devices, or select Devices to select all devices in the list. 4. Indicate whether the endpoints are primary or secondary. 5. Select OK.
Configuring the NDMP device TapeServer group The VTL TapeServer group holds tape drives that interface with NDMP (Network Data Management Protocol)-based backup applications and that send control information and data streams over IP (Internet Protocol) instead of Fibre Channel (FC). A device used by the NDMP TapeServer must be in the VTL group TapeServer and is available only to the NDMP TapeServer.
324 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Procedure 1. Add tape drives to a new or existing library (in this example, named “dd990-16”). 2. Create slots and CAPs for the library. 3. Add the created devices in a library (in this example, “dd990-16”) to the TapeServer access group. 4. Enable the NDMP daemon by entering at the command line: # ndmpd enable Starting NDMP daemon, please wait. NDMP daemon is enabled. 5. Ensure that the NDMP daemon sees the devices in the TapeServer group: # ndmpd show devicenames NDMP Device Virtual Name Vendor Product Serial Number ------/dev/dd_ch_c0t0l0 dd990-16 changer STK L180 6290820000 /dev/dd_st_c0t1l0 dd990-16 drive 1 IBM ULTRIUM-TD3 6290820001 /dev/dd_st_c0t2l0 dd990-16 drive 2 IBM ULTRIUM-TD3 6290820002 /dev/dd_st_c0t3l0 dd990-16 drive 3 IBM ULTRIUM-TD3 6290820003 /dev/dd_st_c0t4l0 dd990-16 drive 4 IBM ULTRIUM-TD3 6290820004 ------6. Add an NDMP user (ndmp in this example) with the following command: # ndmpd user add ndmp Enter password: Verify password: 7. Verify that user ndmp is added correctly: # ndmpd user show ndmp 8. Display the NDMP configuration: # ndmpd option show all Name Value ------authentication text debug disabled port 10000 preferred-ip ------9. Change the default user password authentication to use MD5 encryption for enhanced security, and verify the change (notice the authentication value changed from text to md5): # ndmpd option set authentication md5 # ndmpd option show all Name Value ------authentication md5 debug disabled port 10000 preferred-ip ------Results NDMP is now configured, and the TapeServer access group shows the device configuration. See the ndmpd chapter of the EMC Data Domain Operating System Command Reference Guide for the complete command set and options.
Working with resources Selecting Resources > Resources displays information about initiators and endpoints. An initiator is a backup client that connects to a system to read and write data using the
Working with resources 325 DD Virtual Tape Library
Fibre Channel (FC) protocol. A specific initiator can support DD Boost over FC or VTL, but not both. An endpoint is the logical target on a DD system to which the initiator connects.
Table 146 Initiators tab
Name Name of initiator, which is either the WWPN or the alias assigned to the initiator. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. Online Endpoints Group name where ports are seen by initiator. Displays None or Offline if the initiator is unavailable.
Table 147 Endpoints tab
Name Specific name of endpoint. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. System Address System address for the endpoint. Enabled HBA (host bus adapter) port operational state, which is either Yes (enabled) or No (not enabled). Status DD system VTL link status, which is either Online (capable of handling traffic) or Offline.
Configure Resources Selecting Configure Resources takes you to the Fibre Channel area, where you can configure endpoints and initiators.
Working with initiators Selecting Resources > Resources > Initiators displays information about initiators. An initiator is a client system FC HBA (fibre channel host bus adapter) WWPN (worldwide port name) with which the DD system interfaces. An initiator name is an alias for the client’s WWPN, for ease of use. While a client is mapped as an initiator – but before an access group has been added – the client cannot access any data on a DD system. After adding an access group for the initiator or client, the client can access only the devices in that access group. A client can have access groups for multiple devices.
326 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
An access group may contain multiple initiators, but an initiator can exist in only one access group.
A maximum of 1024 initiators can be configured for a DD system.
Table 148 Initiator information
Name Name of initiator. Group Group associated with initiator. Online Endpoints Endpoints seen by initiator. Displays none or offline if initiator is unavailable. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. Vendor Name Name of vendor for initiator.
Selecting Configure Initiators takes you to the Fibre Channel area, where you can configure endpoints and initiators. CLI Equivalent # vtl initiator show Initiator Group Status WWNN WWPN Port ------tsm6_p1 tsm3500_a Online 20:00:00:24:ff:31:ce:f8 21:00:00:24:ff:31:ce:f8 10b ------
Initiator Symbolic Port Name Address Method ------tsm6_p1 QLE2562 FW:v5.06.03 DVR:v8.03.07.15.05.09-k auto ------
Working with endpoints Selecting Resources > Resources > Endpoints provides information about endpoint hardware and connectivity.
Table 149 Hardware Tab
System Address System address of endpoint. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node.
Working with endpoints 327 DD Virtual Tape Library
Table 149 Hardware Tab (continued)
Enabled HBA (host bus adapter) port operational state, which is either Yes (enabled) or No (not enabled).
NPIV NPIV status of this endpoint: eithe Enabled or Disabled. LInk Status Link status of this endpoint: either Online or Offline. Operation Status Operation status of this endpoint: either Normal or Marginal. # of Endpoints Number of endpoints associated with this endpoint.
Table 150 Endpoints Tab
Name Specific name of endpoint. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel (FC) port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. System Address System address of endpoint. Enabled HBA (host bus adapter) port operational state, which is either Yes (enabled) or No (not enabled).
LInk Status Link status of this endpoint: either Online or Offline.
Configure Endpoints Selecting Configure Endpoints takes you to the Fibre Channel area, where you can change any of the above information for the endpoint. CLI Equivalent # scsitarget endpoint show list Endpoint System Address Transport Enabled Status ------endpoint-fc-0 5a FibreChannel Yes Online endpoint-fc-1 5b FibreChannel Yes Online
Working with a selected endpoint Selecting Resources > Resources > Endpoints > endpoint provides information about the endpoint's hardware, connectivity, and statistics.
Table 151 Hardware tab
System Address System address of endpoint.
328 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Table 151 Hardware tab (continued)
WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. Enabled HBA (host bus adapter) port operational state, which is either Yes (enabled) or No (not enabled).
NPIV NPIV status of this endpoint: eithe Enabled or Disabled. LInk Status Link status of this endpoint: either Online or Offline. Operation Status Operation status of this endpoint: either Normal or Marginal. # of Endpoints Number of endpoints associated with this endpoint.
Table 152 Summary tab
Name Specific name of endpoint. WWPN Unique worldwide port name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the Fibre Channel port. WWNN Unique worldwide node name, which is a 64-bit identifier (a 60- bit value preceded by a 4-bit Network Address Authority identifier), of the FC node. System Address System address of endpoint. Enabled HBA (host bus adapter) port operational state, which is either Yes (enabled) or No (not enabled).
LInk Status Link status of this endpoint: either Online or Offline.
Table 153 Statistics tab
Endpoint Specific name of endpoint. Library Name of library containing the endpoint. Device Number of device. Ops/s Operations per second. Read KiB/s Speed of reads in KiB per second. Write KiB/s Speed of writes in KiB per second.
Working with a selected endpoint 329 DD Virtual Tape Library
Table 154 Detailed Statistics tab
Endpoint Specific name of endpoint. # of Control Commands Number of control commands. # of Read Commands Number of read commands. # of Write Commands Number of write commands. In (MiB) Number of MiB written (the binary equivalent of MB). Out (MiB) Number of MiB read. # of Error Protocol Number of error protocols. # of Link Fail Number of link failures. # of Invalid Crc Number of invalid CRCs (cyclic redundancy checks).
# of Invalid TxWord Number of invalid tx (transmission) words. # of Lip Number of LIPs (loop initialization primitives). # of Loss Signal Number of signals or connections that have been lost. # of Loss Sync Number of signals or connections that have lost synchronization.
Working with pools Selecting Pools > Pools displays detailed information for the Default pool and any other existing pools. A pool is a collection of tapes that maps to a directory on the file system. Pools are used to replicate tapes to a destination. You can convert directory-based pools to MTree-based pools to take advantage of the greater functionality of MTrees. Note the following about pools:
l Pools can be of two types: MTree (recommended), or Directory, which is backward- compatible. l A pool can be replicated no matter where individual tapes are located. Tapes can be in the vault or in a library (slot, cap, or drive). l You can copy and move tapes from one pool to another. l Pools are not accessible by backup software. l No VTL configuration or license is needed on a replication destination when replicating pools. l You must create tapes with unique barcodes. Duplicate barcodes may cause unpredictable behavior in backup applications and can be confusing to users. l Two tapes in two different pools on a DD system may have the same name, and in this case, neither tape can be moved to the other tape's pool. Likewise, a pool sent to a replication destination must have a name that is unique on the destination.
330 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Table 155 Pools tab
Location The location of the pool. Type Whether it is a Directory or MTree pool. Tape Count The number of tapes in the pool. Capacity The total configured data capacity of tapes in the pool, in GiB (Gibibytes base-2 equivalent of GB, Gigabytes). Used The amount of space used on virtual tapes in the pool. Average Compression The average amount of compression achieved for data on tapes in the pool.
Table 156 Replication tab
Name The name of the pool. Configured Whether replication is configured for the pool: yes or no. Remote Source Contains an entry only if the pool is replicated from another DD system. Remote Destination Contains an entry only if the pool replicates to another DD system.
From the More Tasks menu, you can create and delete pools, as well as search for tapes.
Creating pools You can create backward-compatible pools, if necessary for your setup, for example, for replication with a pre-5.2 DD OS system. Procedure
1. Select Pools > Pools. 2. Select More Tasks > Pool > Create. 3. In the Create Pool dialog, enter a Pool Name, noting that a pool name:
l cannot be “all,” “vault,” or “summary.” l cannot have a space or period at its beginning or end. l is case-sensitive.
4. If you want to create a directory pool (which is backward compatible with the previous version of DD System Manager), select the option “Create a directory backwards compatibility mode pool. ” However, be aware that the advantages of using an MTree pool include the ability to:
l make individual snapshots and schedule snapshots. l apply retention locks. l set an individual retention policy.
Creating pools 331 DD Virtual Tape Library
l get compression information. l get data migration policies to the Retention Tier. l establish a storage space usage policy (quota support) by setting hard limits and soft limits.
5. Select OK to display the Create Pool Status dialog. 6. When the Create Pool Status dialog shows Completed, select Close. The pool is added to the Pools subtree, and you can now add virtual tapes to it.
# vtl pool add VTL_Pool A VTL pool named VTL_Pool is added.
Deleting pools Before a pool can be deleted, you must have deleted any tapes contained within it. If replication is configured for the pool, the replication pair must also be deleted. Deleting a pool corresponds to renaming the MTree and then deleting it, which occurs at the next cleaning process. Procedure
1. Select Pools > Pools pool. 2. Select More Tasks > Pool > Delete. 3. In the Delete Pools dialog, select the checkbox of items to delete:
l The name of each pool, or l Pool Names, to delete all pools.
4. Select Submit in the confirmation dialogs. 5. When the Delete Pool Status dialog shows Completed, select Close. The pool will have been removed from the Pools subtree.
Working with a selected pool Both Virtual Tape Libraries > VTL Service > Vault > pool and Pools > Pools > pool display detailed information for a selected pool. Notice that pool “Default” always exists.
Table 157 Pool Tab
Convert to MTree Pool Select this button to convert a Directory pool to an MTree pool. Type Whether it is a Directory or MTree pool. Tape Count The number of tapes in the pool. Capacity The total configured data capacity of tapes in the pool, in GiB (Gibibytes, base-2 equivalent of GB, Gigabytes). Used The amount of space used on virtual tapes in the pool.
332 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Table 157 Pool Tab (continued)
Average Compression The average amount of compression achieved for data on tapes in the pool.
Table 158 Replication Tab
Name The name of the pool. Configured Whether replication is configured for this pool: yes or no. Remote Source Contains an entry only if the pool is replicated from another DD system.
Remote Destination Contains an entry only if the pool replicates to another DD system.
You can also select the Replication Detail button, at the top right, to go directly to the Replication information panel for the selected pool. From either the Virtual Tape Libraries or Pools area, from the More Tasks menu, you can create, delete, move, copy, or search for a tape in the pool. From the Pools area, from the More Tasks menu, you can rename or delete a pool.
Converting a directory pool to an MTree pool MTree pools have many advantages over directory pools. See the Creating pools section for more information. Procedure 1. Make sure the following prerequisites have been met: l The source and destination pools must have been synchronized, so that the number of tapes, and the data on each side, remains intact. l The directory pool must not be a replication source or destination. l The file system must not be full. l The file system must not have reached the maximum number of MTrees allowed (100). l There must not already be an MTree with the same name. l If the directory pool is being replicated on multiple systems, those replicating systems must be known to the managing system. l If the directory pool is being replicated to an older DD OS (for example, from DD OS 5.5 to DD OS 5.4), it cannot be converted. As a workaround: n Replicate the directory pool to a second DD system. n Replicate the directory pool from the second DD system to a third DD system. n Remove the second and third DD systems from the managing DD system's Data Domain network. n On any of the systems running DD OS 5.5, from the Pools submenu, select Pools and a directory pool. In the Pools tab, select Convert to MTree Pool.
Converting a directory pool to an MTree pool 333 DD Virtual Tape Library
2. With the directory pool you wish to convert highlighted, choose Convert to MTree Pool. 3. Select OK in the Convert to MTree Pool dialog. 4. Be aware that conversion affects replication in the following ways:
l VTL is temporarily disabled on the replicated systems during conversion. l The destination data is copied to a new pool on the destination system to preserve the data until the new replication is initialized and synced. Afterward, you may safely delete this temporarily copied pool, which is named CONVERTED-pool, where pool is the name of the pool that was upgraded (or the first 18 characters for long pool names). [This applies only to DD OS 5.4.1.0 and later.] l The target replication directory will be converted to MTree format. [This applies only to DD OS 5.2 and later.] l Replication pairs are broken before pool conversion and re-established afterward if no errors occur. l Retention Lock cannot be enabled on systems involved in MTree pool conversion.
Moving tapes between pools If they reside in the vault, tapes can be moved between pools to accommodate replication activities. For example, pools are needed if all tapes were created in the Default pool, but you later need independent groups for replicating groups of tapes. You can create named pools and re-organize the groups of tapes into new pools.
You cannot move tapes from a tape pool that is a directory replication source. As a workaround, you can:
l Copy the tape to a new pool, then delete the tape from the old pool. l Use an MTree pool, which allows you to move tapes from a tape pool that is a directory replication source.
1. With a pool highlighted, select More Tasks > Tapes > Move. Note that when started from a pool, the Tapes Panel allows tapes to be moved only between pools.
2. In the Move Tapes dialog, enter information to search for the tapes to move, and select Search:
Table 159 Move Tapes dialog
Field User input
Location Location cannot be changed. Pool Select the name of the pool where the tapes reside. If no pools have been created, use the Default pool. Barcode Specify a unique barcode. or leave the default (*) to import a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters.
334 EMC Data Domain Operating System 5.7 Administration Guide DD Virtual Tape Library
Table 159 Move Tapes dialog (continued)
Field User input
Count Enter the maximum number of tapes you want to be returned to you. If you leave this blank, the barcode default (*) is used. Tapes Per Select the maximum number of tapes to display per page. Possible values are 15, Page 30, and 45. Items Shows the number of tapes selected across multiple pages – updated Selected automatically for each tape selection.
3. From the search results list, select the tapes to move. 4. From the Select Destination: Location list, select the location of the pool to which tapes are to be moved. This option is available only when started from the (named) Pool view. 5. Select Next. 6. From the Move Tapes view, verify the summary information and tape list, and select Submit. 7. Select Close in the status window.
Copying tapes between pools Tapes can be copied between pools, or from the vault to a pool, to accommodate replication activities. This option is available only when started from the (named) Pool view. Procedure
1. With a pool highlighted, select More Tasks > Tapes > Copy. 2. In the Copy Tapes Between Pools dialog. select the checkboxes of tapes to copy, or enter information to search for the tapes to copy, and select Search:
Table 160 Copy Tapes Between Pools dialog
Field User input
Location Select either a library or the Vault for locating the tape. While tapes always show up in a pool (under the Pools menu), they are technically in either a library or the vault, but not both, and they are never in two libraries at the same time. Use the import/export options to move tapes between the vault and a library. Pool To copy tapes between pools, select the name of the pool where the tapes currently reside. If no pools have been created, use the Default pool. Barcode Specify a unique barcode. or leave the default (*) to import a group of tapes. Barcode allows the wildcards ? and *, where ? matches any single character and * matches 0 or more characters. Count Enter the maximum number of tapes you want to be imported. If you leave this blank, the barcode default (*) is used. Tapes Per Select the maximum number of tapes to display per page. Possible values are 15, Page 30, and 45.
Copying tapes between pools 335 DD Virtual Tape Library
Table 160 Copy Tapes Between Pools dialog (continued)
Field User input
Items Shows the number of tapes selected across multiple pages – updated Selected automatically for each tape selection.
3. From the search results list, select the tapes to copy. 4. From the Select Destination: Pool list, select the pool where tapes are to be copied. If a tape with a matching barcode already resides in the destination pool, an error is displayed, and the copy aborts. 5. Select Next. 6. From the Copy Tapes Between Pools dialog, verify the summary information and the tape list, and select Submit. 7. Select Close on the Copy Tapes Between Pools Status window.
Renaming pools A pool can be renamed only if none of its tapes is in a library. Procedure
1. Select Pools > Pools pool. 2. Select More Tasks > Pool > Rename. 3. In the Rename Pool dialog, enter the new Pool Name, with the caveat that this name:
l cannot be “all,” “vault,” or “summary.” l cannot have a space or period at its beginning or end. l is case-sensitive.
4. Select OK to display the Rename Pool status dialog. 5. After the Rename Pool status dialog shows Completed, select OK. The pool will have been renamed in the Pools subtree in both the Pools and the Virtual Tape Libraries areas.
336 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 14 DD Replicator
This chapter includes:
l DD Replicator overview. 338 l Replication between HA and non-HA systems. 338 l Prerequisites for replication configuration. 339 l Replication version compatibility. 340 l Using DD Encryption with DD Replicator. 343 l Replication types. 344 l Replication topologies. 348 l Managing replication. 353 l Monitoring replication . 369
DD Replicator 337 DD Replicator
DD Replicator overview EMC Data Domain Replicator (DD Replicator) provides automated, policy-based, network- efficient, and encrypted replication for DR (disaster recovery) and multi-site backup and archive consolidation. DD Replicator asynchronously replicates only compressed, deduplicated data over a WAN (wide area network). DD Replicator performs two levels of deduplication to significantly reduce bandwidth requirements: local and cross-site deduplication. Local deduplication determines the unique segments to be replicated over a WAN. Cross-site deduplication further reduces bandwidth requirements when multiple sites are replicating to the same destination system. With cross-site deduplication, any redundant segment previously transferred by any other site, or as a result of a local backup or archive, will not be replicated again. This improves network efficiency across all sites and reduces daily network bandwidth requirements up to 99%, making network-based replication fast, reliable, and cost- effective. In order to meet a broad set of DR requirements, DD Replicator provides flexible replication topologies, such as full system mirroring, bi-directional, many-to-one, one-to- many, and cascaded. In addition, you can choose to replicate either all or a subset of the data on your DD system. For the highest level of security, DD Replicator can encrypt data being replicated between DD systems using the standard SSL (Secure Socket Layer) protocol. DD Replicator scales performance and supported fan-in ratios to support large enterprise environments. When deployed over a 10GB network, DD Replicator can mirror data between two systems at up to 52 TB/hour. Before getting started with DD Replicator, note the following general requirements:
l DD Replicator is a licensed product. See your EMC Data Domain sales representative to purchase licenses. l You can usually replicate only between machines that are within two releases of each other, for example, from 5.5 to 5.7. However, there may be exceptions to this (as a result of atypical release numbering), so review the tables in the Replication version compatibility section, or check with your EMC representative. l If you are unable to manage and monitor DD Replicator from the current version of the DD System Manager, use the replication commands described in the EMC Data Domain Operating System Command Reference Guide.
Replication between HA and non-HA systems If you want to set up a replication between a high-availability (HA) system and a system running DD OS 5.7.0.3 or earlier, you must create and manage that replication on the HA system if you want to use the DD System Manager graphical user interface (GUI). However, you can perform replications from a non-HA system to an HA system using the CLI as well as from the HA system to the non-HA system.
338 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
Prerequisites for replication configuration Before configuring a replication, review the following prerequisites to minimize initial data transfer time, prevent overwriting of data, etc.
l Contexts – Determine the maximum number of contexts for your DD systems by reviewing the replication streams numbers in the following table. Table 161 Data streams sent to a Data Domain system
Model RAM/ NVRAM Backup Backup Repla Repla dest Mixed write read source streams streams streams streams
DD140, DD160, 4 GB or 6 GB / 16 4 15 20 w SN) can be used for all replication types. l Extended Retention (ER)-enabled systems cannot be the source for directory replication. l Collection replication cannot be configured from either a single node (SN) system to an Extended Retention (ER)-enabled system, nor from an ER-enabled system to an SN system. In this table:
l SN = single node DD system (no Extended Retention) l ER = Extended Retention-enabled DD system
Table 165 Topology Support by Replication Type and DD System Type
Topologies MTree Replication Directory Replication Collection Replication
one-to-one -> SN -> SN SN -> SN ER->SN [supported starting SN -> ER ER -> ER with 5.5 release; prior to 5.5, it is recovery only]
one-to-one -> SN -> SN not supported bidirectional
one-to-many -> SN -> SN not supported SN -> ER
many-to-one -> SN -> SN not supported SN -> ER
cascaded -> -> SN -> SN ER -> ER -> ER ER> SN -> SN -> ER SN -> SN -> SN
Cascaded replication supports mixed topologies where the second leg in a cascaded connection is different from the first type in a connection (for example, A -> B is directory replication, and B -> C is collection replication).
348 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
Table 166 Mixed Topologies Supported with Cascaded Replication
SN – Dir Repl -> ER – MTree Repl -> ER – MTree SN – Dir Repl -> ER – Col Repl -> ER – Col Repl Repl
SN – MTree Repl -> SN – Col Repl -> SN – Col SN – MTree Repl -> ER – Col Repl -> ER – Col Repl Repl
One-to-one replication The simplest type of replication is from a DD source system to a DD destination system, otherwise known as a one-to-one replication pair. This replication topology can be configured with directory, MTree, or collection replication types. Figure 21 One-to-one replication pair
One-to-one replication 349 DD Replicator
Bi-directional replication In a bi-directional replication pair, data from a directory or MTree on DD system A is replicated to DD system B, and from another directory or MTree on DD system B to DD system A. Figure 22 Bi-directional replication
350 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
One-to-many replication In one-to-many replication, data flows from a source directory or MTree on one DD system to several destination DD systems. You could use this type of replication to create more than two copies for increased data protection, or to distribute data for multi-site usage. Figure 23 One-to-many replication
Many-to-one replication In many-to-one replication, whether with MTree or directory, replication data flows from several source DD systems to a single destination DD system. This type of replication can
One-to-many replication 351 DD Replicator
be used to provide data recovery protection for several branch offices on a corporate headquarter’s IT system. Figure 24 Many-to-one replication
Cascaded replication In a cascaded replication topology, a source directory or MTree is chained among three DD systems. The last hop in the chain can be configured as collection, MTree, or directory replication, depending on whether the source is directory or MTree. For example, DD system A replicates one or more MTrees to DD system B, which then replicates those MTrees to DD system C. The MTrees on DD system B are both a destination (from DD system A) and a source (to DD system C).
352 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
Figure 25 Cascaded directory replication
Data recovery can be performed from the non-degraded replication pair context. For example:
l In the event DD system A requires recovery, data can be recovered from DD system B. l In the event DD system B requires recovery, the simplest method is to perform a replication resync from DD system A to (the replacement) DD system B. In this case, the replication context from DD system B to DD system C should be broken first. After the DD system A to DD system B replication context finishes resync, a new DD system B to DD System C context should be configured and resynced.
Managing replication You can manage replication using the Data Domain System Manager (DD System Manager) or the Data Domain Operating System (DD OS) Command Line Interface (CLI). To use a graphical user interface (GUI) to manage replication, log in to the DD System Manager. Procedure 1. From the menu at the left of the DD System Manager, select Replication. If your license has not been added yet, select Add License. 2. Select Automatic or On-Demand (you must have a DD Boost license for on-demand). CLI Equivalent You can also log in at the CLI: login as: sysadmin Data Domain OS 5.5.x.x-12345
Managing replication 353 DD Replicator
Using keyboard-interactive authentication. Password:
Replication status Replication Status shows the system-wide count of replication contexts exhibiting a warning (yellow text) or error (red text) state, or if conditions are normal.
Summary view The Summary view lists the configured replication contexts for a DD system, displaying aggregated information about the selected DD system – that is, summary information about the inbound and outbound replication pairs. The focus is the DD system, itself, and the inputs to it and outputs from it. The Summary table can be filtered by entering a Source or Destination name, or by selecting a State (Error, Warning, or Normal).
Table 167 Replication Summary view
Source System and path name of the source context, with format system.path. For example, for directory dir1 on system dd120-22, you would see dd120-22.chaos.local/data/ col1/dir1.
Destination System and path name of destination context, with format system.path. For example, for MTree MTree1 on system dd120-44, you would see dd120-44.chaos.local/data/ col1/MTree1.
Type Type of context: MTree, directory (Dir), or Pool. State Possible states of replication pair status include:
l Normal – If the replica is Initializing, Replicating, Recovering, Resyncing, or Migrating.
l Idle – For MTree replication, this state can display if the replication process is not currently active or for network errors (such as the destination system being inaccessible).
l Warning – If there is an unusual delay for the first five states, or for the Uninitialized state.
l Error – Any possible error states, such as Disconnected.
Synced As Of Time Timestamp for last automatic replication sync operation performed by the source. For MTree replication, this value is updated when a snapshot is exposed on the destination. For directory replication, it is updated when a sync point inserted by the source is applied. A value of unknown displays during replication initialization. Pre-Comp Remaining Amount of pre-compressed data remaining to be replicated. Completion Time (Est.) Value is either Completed, or the estimated amount of time required to complete the replication data transfer based on the last 24 hours’ transfer rate.
354 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
Detailed information for a replication context Selecting one replication context from the Summary view populates that context’s information in Detailed Information, Performance Graph, Completion Stats, and Completion Predictor.
Table 168 Detailed Information
State Description Message about state of replica. Source System and path name of source context, with format system.path. For example, for directory dir1 on system dd120-22, you would see dd120-22.chaos.local/data/ col1/dir1.
Destination System and path name of destination context, with format system.path. For example, for MTree MTree1 on system dd120-44, you would see dd120-44.chaos.local/data/ col1/MTree1.
Connection Port System name and listen port used for replication connection.
Table 169 Performance Graph
Pre-Comp Remaining Pre-compressed data remaining to be replicated. Pre-Comp Written Pre-compressed data written on the source. Post-Comp Replicated Post-compressed data that has been replicated.
Table 170 Completion Stats
Synced As Of Time Timestamp for last automatic replication sync operation performed by the source. For MTree replication, this value is updated when a snapshot is exposed on the destination. For directory replication, it is updated when a sync point inserted by the source is applied. A value of unknown displays during replication initialization. Completion Time (Est.) Value is either Completed or the estimated amount of time required to complete the replication data transfer based on the last 24 hours’ transfer rate. Pre-Comp Remaining Amount of data remaining to be replicated. Files Remaining (Directory Replication Only) Number of files that have not yet been replicated. Status For source and destination endpoints, shows status (Enabled, Disabled, Not Licensed, etc.) of major components on the system, such as:
Summary view 355 DD Replicator
Table 170 Completion Stats (continued)
l Replication Lock
l Encryption at Rest
l Encryption over Wire
l Available Space
l Low Bandwidth Optimization
l Compression Ratio
l Low Bandwidth Optimization Ratio
Completion Predictor The Completion Predictor is a widget for tracking a backup job's progress and for predicting when replication will complete, for a selected context.
Creating a replication pair Before creating a replication pair, make sure the destination does not exist, or you will get an error. Procedure
1. Select Replication > Automatic > Summary tab > Create Pair . 2. In the Create Pair dialog, add information to create an inbound or outbound MTree, directory, collection, or pool replication pair, as described in the next sections. Adding a DD system for replication You may need to add a DD system as either a host or a destination before you can create a replication pair.
Make sure the system being added is running a compatible DD OS version.
Procedure 1. In the Create Pair dialog, select Add System. 2. For System, enter the hostname or IP address of the system to be added. 3. For User Name and Password, enter the sysadmin's user name and password. 4. Optionally, select More Options to enter a proxy IP address (or system name) of a system that cannot be reached directly. If configured, enter a custom port instead of the default port 3009.
IPv6 addresses are supported only when adding a DD OS 5.5 or later system to a management system using DD OS 5.5 or later.
356 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
If the system is unreachable after adding it to DD System Manager, make sure that there is a route from the managing system to the system being added. If a hostname (either a fully qualified domain name (FQDN) or non-FQDN) is entered, make sure it is resolvable on the managed system. Configure a domain name for the managed system, ensure a DNS entry for the system exists, or ensure an IP address to hostname mapping is defined.
6. If the system certificate is not verified, the Verify Certificate dialog shows details about the certificate. Check the system credentials. Select OK if you trust the certificate, or select Cancel. Creating a collection replication pair See the Collection replication section for general information about this type of replication. Before creating a collection replication pair, make sure: l The storage capacity of the destination system is equal to, or greater than, that of the source system. (If the destination capacity is less than that of the source, the available capacity on the source is reduced to that of the destination.) l The destination has been destroyed, and subsequently re-created, but not enabled. l Each destination and each source is in only one context at a time. l The file system is disabled on the replica, while configuring and enabling encryption on the source. l The file system is disabled on the source, while configuring and enabling encryption on the replica. Procedure 1. In the Create Pair dialog, select Collection from the Replication Type menu. 2. Select the source system hostname from the Source System menu. 3. Select the destination system hostname from the Destination System menu. The list includes only those hosts in the DD-Network list. 4. If you want to change any host connection settings, select the Advanced tab. 5. Select OK. Replication from the source to the destination begins. Results Test results from Data Domain returned the following performance guidelines for replication initialization. These are guidelines only, and actual performance seen in production environments may vary. l Over a gibibit LAN: With a high enough shelf count to drive maximum input/output and ideal conditions, collection replication can saturate a 1GigE link (modulo 10% protocol overhead), as well as 400-900 MB/sec on 10gigE, depending on the platform. l Over a WAN, performance is governed by the WAN link line speed, bandwidth, latency, and packet loss rate.
Summary view 357 DD Replicator
Creating an MTree, directory, or pool replication pair See the MTree replication and Directory replication sections for general information about these types of replication. When creating an MTree, directory, or pool replication pair:
l Make sure the replication is transiting\exiting the correct interface. When defining a replication context, the host names of the source and destination must resolve with forward and reverse lookups. To make the data transit alternate interfaces on the system, other than the default resolving interface, the replication context must be modified after creation. It may be necessary to set up host files to ensure that contexts are defined on non-resolving (cross-over) interfaces. l You can “reverse” the context for an MTree replication, that is, you can switch the destination and the source. l Subdirectories within an MTree cannot be replicated, because the MTree, in its entirety, is replicated. l MTree replication is supported from Extended Retention-enabled systems to non- Extended Retention-enabled systems, if both are running DD OS 5.5 or later. l The destination DD system must have available storage capacity of at least the post- compressed size of the expected maximum post-compressed size of the source directory or MTree. l When replication is initialized, a destination directory is created automatically. l A DD system can simultaneously be the source for one context and the destination for another context. Procedure 1. In the Create Pair dialog, select Directory, MTree (default), or Pool from the Replication Type menu. 2. Select the source system hostname from the Source System menu. 3. Select the destination system hostname from the Destination System menu. 4. Enter the source path in the Source Path text box (notice the first part of the path is a constant that changes based on the type of replication chosen). 5. Enter the destination path in the Destination Directory text box (notice the first part of the path is a constant that changes based on the type of replication chosen). 6. If you want to change any host connection settings, select the Advanced tab. 7. Select OK. The Replication from the source to the destination begins. Test results from Data Domain returned the following guidelines for estimating the time needed for replication initialization. These are guidelines only and may not be accurate in specific production environments.
l Using a T3 connection, 100ms WAN, performance is about 40 MiB/sec of pre- compressed data, which gives data transfer of: 40 MiB/sec = 25 seconds/GiB = 3.456 TiB/day l Using the base-2 equivalent of gigabit LAN, performance is about 80 MiB/sec of pre-compressed data, which gives data transfer of about double the rate for a T3 WAN.
358 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
CLI Equivalent Here are examples of creating MTree and directory replication pairs at the CLI. The last example specifies the IP version used as a replication transport. # replication add source mtree://ddsource.test.com/data/col1/ examplemtree destination mtree://ddtarget.test.com/data/col1/ examplemtree (Mtree example) # replication add source dir://ddsource.test.com/data/col1/ directorytorep destination dir://ddtarget.test.com/backup/ directorytorep # replication add source dir://ddsource.test.com/data/col1/ directorytorep destination dir://ddtarget.test.com/backup/ directorytorep ipversion ipv6 To start replication between a source and destination, use the replication initialize command on the source. This command checks that the configuration and connections are correct and returns error messages if any problems appear. # replication initialize mtree://host3.test.com/data/col1/mtree1/
Configuring bi-directional replication To create a bi-directional replication pair, use the directory or MTree replication pair procedure (for example, using mtree2) from host A to host B. Use the same procedure to create a replication pair (for example, using mtree1) from host B to host A. For this configuration, destination pathnames cannot be the same. Configuring one-to-many replication To create a one-to-many replication pair, use the directory or MTree replication pair procedure (for example, using mtree1) on host A to: (1) mtree1 on host B, (2) mtree1 on host C, and (3) mtree1 on host D. A replication recovery cannot be done to a source context whose path is the source path for other contexts; the other contexts must be broken and resynced after the recovery. Configuring many-to-one replication To create a many-to-one replication pair, use the directory or MTree replication pair procedure [for example, (1) mtree1 from host A to mtree1 on host C and (2) mtree2 on host B to mtree2 on host C.] Configuring cascaded replication To create a cascaded replication pair, use the directory or MTree replication pair procedure: (1) mtree1 on host A to mtree1 on host B, and (2) on host B, create a pair for mtree1 to mtree1 on host C. The final destination context (on host C in this example, but more than three hops are supported) can be a collection replica or a directory or MTree replica.
Disabling and enabling a replication pair Disabling a replication pair temporarily pauses the active replication of data between a source and a destination. The source stops sending data to the destination, and the destination stops serving as an active connection to the source. Procedure 1. Select one or more replication pairs in the Summary table, and select Disable Pair. 2. In the Display Pair dialog, select Next and then OK. 3. To resume operation of a disabled replication pair, select one or more replication pairs in the Summary table, and select Enable Pair to display the Enable Pair dialog. 4. Select Next and then OK. Replication of data is resumed.
Summary view 359 DD Replicator
Deleting a replication pair When a directory or MTree replication pair is deleted, the destination directory or MTree, respectively, becomes writeable. When a collection replication pair is deleted, the destination DD system becomes a stand-alone read/write system, and the file system is disabled. Procedure 1. Select one or more replication pairs in the Summary table, and select Delete Pair. 2. In the Delete Pair dialog, select Next and then OK. The replication pairs are deleted.
CLI Equivalent Before running this command, always run the filesys disable command. Then, afterward, run the filesys enable command # replication break
Changing host connection settings To direct traffic out of a specific port, modify a current context by altering the connection host parameter using a host name previously defined in the local hosts file to address the alternate system. That host name will correspond to the destination. The host entry will indicate an alternate destination address for that host. This may be required on both the source and destination systems. Procedure 1. Select the replication pair in the Summary table, and select Modify Settings. You can also change these settings when you are performing Create Pair, Start Resync, or Start Recover by selecting the Advanced tab. 2. In the Modify Connection Settings dialog, modify any or all of these settings: a. Use Low Bandwidth Optimization – For enterprises with small data sets and 6 Mb/s or less bandwidth networks, DD Replicator can further reduce the amount of data to be sent using low bandwidth optimization. This enables remote sites with limited bandwidth to use less bandwidth or to replicate and protect more of their data over existing networks. Low bandwidth optimization must be enabled on both the source and destination DD systems. If the source and destination have incompatible low bandwidth optimization settings, low bandwidth optimization will be inactive for that context. After enabling low bandwidth optimization on the source and destination, both systems must undergo a full cleaning cycle to prepare the existing data, so run filesys clean start on both systems. The duration of the cleaning cycle depends on the amount of data on the DD system, but takes longer than a normal cleaning. For more information on the filesys commands, see the EMC Data Domain Operating System Command Reference Guide. Important: Low bandwidth optimization is not supported if the DD Extended Retention software option is enabled on either DD system. It is also not supported for Collection Replication.
b. Enable Encryption Over Wire – DD Replicator supports encryption of data-in-flight by using standard SSL (Secure Socket Layer) protocol version 3, which uses the ADH-AES256-SHA cipher suite to establish secure replication connections. Both sides of the connection must enable this feature for encryption to proceed.
360 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
c. Network Preference – You may choose IPv4 or IPv6. An IPv6-enabled replication service can still accept connections from an IPv4 replication client if the service is reachable via IPv4. An IPv6-enabled replication client can still communicate with an IPv4 replication service if the service is reachable via IPv4. d. Use Non-default Connection Host – The source system transmits data to a destination system listen port. Since a source system can have replication configured for many destination systems (each of which can have a different listen port), each context on the source can configure the connection port to the corresponding listen port of the destination. 3. Select Next and then Close. The replication pair settings are updated, and replication resumes.
CLI Equivalent #replication modify connection-host [port
Managing replication systems You can add or delete Data Domain systems to be used for replication using the Manage Systems dialog. Procedure 1. Select Modify Settings. 2. In the Manage Systems dialog, add and/or delete Data Domain systems, as required. 3. Select Close.
Recovering data from a replication pair If source replication data becomes inaccessible, it can be recovered from the replication pair destination. The source must be empty before recovery can proceed. Recovery can be performed for all replication topologies, except for MTree replication. Recovery of data from a directory pool, as well as from directory and collection replication pairs, is described in the next sections. Recovering directory pool data You can recover data from a directory-based pool, but not from an MTree-based pool. Procedure
1. Select More > Start Recover. 2. In the Start Recover dialog, select Pool from the Replication Type menu. 3. Select the source system hostname from the System to recover to menu. 4. Select the destination system hostname from the System to recover from menu. 5. Select the context on the destination from which data is recovered. 6. If you want to change any host connection settings, select the Advanced tab. 7. Select OK to start the recovery.
Summary view 361 DD Replicator
Recovering collection replication pair data To successfully recover collection replication pair data, the source file system must be in a pristine state, and the destination context must be fully initialized. Procedure 1. Select More > Start Recover to display the Start Recover dialog. 2. Select Collection from the Replication Type menu. 3. Select the source system host name from the System to recover to menu. 4. Select the destination system host name from the System to recover from menu. 5. Select the context on the destination from which data is recovered. Only one collection will exist on the destination. 6. To change any host connection settings, select the Advanced tab. 7. Select OK to start the recovery. Recovering directory replication pair data To successfully recover directory replication pair data, the same directory used in the original context must be created (but left empty). Procedure 1. Select More > Start Recover to display the Start Recover dialog. 2. Select Directory from the Replication Type menu. 3. Select the host name of the system to which data needs to be restored from the System to recover to menu. 4. Select the host name of the system that will be the data source from the System to recover from menu. 5. Select the context to restore from the context list. 6. To change any host connection settings, select the Advanced tab. 7. Select OK to start the recovery.
Aborting a replication pair recovery If a replication pair recovery fails or must be terminated, you can stop the replication recovery. Procedure 1. Select the More menu, and select Abort Recover to display the Abort Recover dialog, which shows the contexts currently performing recovery. 2. Select the checkbox of one or more contexts to abort from the list. 3. Select OK. After you finish As soon as possible, you should restart recovery on the source.
Resyncing an MTree, directory, or pool replication pair Resynchronization is the process of recovering (or bringing back into sync) the data between a source and a destination replication pair after a manual break. The replication pair are resynchronized so both endpoints contain the same data. Resynchronization is available for MTree, directory, and pool replication, but not for collection replication. A replication resynchronization can also be used:
362 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
l To recreate a context that has been deleted. l When a destination runs out of space, but the source still has data to replicate. l To convert a directory replication pair to an MTree replication pair. Procedure 1. Delete the context on both the replication source and replication destination systems. 2. From either the replication source or replication destination system, select More > Start Resync to display the Start Resync dialog. 3. Select the Replication Type to be resynced: Directory, MTree, or Pool. 4. Select the replication source system host name from the Source System menu. 5. Select the replication destination system host name from the Destination System menu. 6. Enter the replication source path in the Source Path text box. 7. Enter the replication destination path in the Destination Path text box. 8. To change any host connection settings, select the Advanced tab. 9. Select OK.
CLI Equivalent # replication resync Destination
Aborting a replication pair resynchronization If a replication pair resynchronization fails or must be terminated, you can stop the resynchronization. Procedure
1. From either the replication source or replication destination system, select More > Abort Resync to display the Abort Resync dialog, which lists all contexts currently performing resynchronization. 2. Select the checkboxes of one or more contexts to abort their resynchronization. 3. Select OK.
Converting a directory replication pair to an MTree When a directory replication pair is converted to an MTree, the directory data is initialized in an MTree, and the directory replication configuration is deleted. The conversion is started with a replication resync that filters all data from the source DD system to the destination DD system. Here are some estimates for filtering performance: [MiB = MibiBytes, base-2 equivalent of Megabytes; GiB = GibiBytes, base-2 equivalent of Gigabytes; TiB = TibiBytes, base-2 equivalent of Terabytes]
l Over a T3, 100ms WAN, filtering performance is about 100 MiB/sec, providing a data transfer rate of 100 MiB/sec = 10 seconds/GiB = 8.6 TiB/day. l Over a gibibit LAN, filtering performance is about 120 MiB/sec, providing a data transfer rate of 120 MiB/sec = 8.3 seconds/GiB = 10.3 TiB/day. Procedure 1. Create a new MTree on both the source and the destination.
Summary view 363 DD Replicator
2. Select the directory replication pair in the Summary table. 3. Select More > Convert to MTree to display the Convert to MTree dialog. 4. Add the new MTree paths to the source and destination text fields. 5. Select OK. A Warning dialog is displayed, indicating the directory data is being initialized in the new MTree, and the old directory replication configuration is being deleted.
6. Select OK. The Convert to MTree Status dialog is displayed, showing the progress of the conversion.
DD Boost view The DD Boost view provides configuration and troubleshooting information to NetBackup administrators who have configured DD systems to use DD Boost AIR (Automatic Image Replication) or any DD Boost application that uses managed file replication. See the EMC Data Domain Boost for OpenStorage Administration Guide for DD Boost AIR configuration instructions. The File Replication tab displays:
l Currently Active File Replication:
n Direction (Out-Going and In-Coming) and the number of files in each. n Remaining data to be replicated (pre-compressed value in GiB) and the amount of data already replicated (pre-compressed value in GiB). n Total size: The amount of data to be replicated and the already replicated data (pre-compressed value in GiB). l Most Recent Status: Total file replications and whether completed or failed
n during the last hour n over the last 24 hours l Remote Systems:
n Select a replication from the list. n Select the time period to be covered from the menu. n Select Show Details for more information about these remote system files. The Storage Unit Associations tab displays the following information, which you can use for audit purposes or to check the status of DD Boost AIR events used for the storage unit's image replications:
l A list of all storage unit Associations known to the system. The source is on the left, and the destination is on the right. This information shows the configuration of AIR on the Data Domain system. l The Event Queue is the pending event list. It shows the local storage unit, the event ID, and the status of the event. An attempt is made to match both ends of a DD Boost path to form a pair and present this as one pair/record. If the match is impossible, for various reasons, the remote path will be listed as Unresolved.
364 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
Remote system files The Show Details button provides information for the selected remote file replication system. File Replications shows starting and ending information, as well as size and data amount, for the selected remote file replication system. The Performance Graph shows performance over time for the selected remote file replication system.
Table 171 File Replications
Start Starting point of time period. End Ending point of time period. File Name Name of specific replication file. Status Most recent status (Success, Failure). Pre-Comp Size (MiB) Amount of pre-compressed outbound and inbound data, as compared to network throughput or post-compressed data (in MiB). Network Bytes (MiB) Amount of network throughput data (in MiB).
Table 172 Performance Graph
Duration Duration for replication (either 1d, 7d or 30d). Interval Interval for replication (either Daily or Weekly). Pre-Comp Replicated Amount of pre-compressed outbound and inbound data (in GiB). Post-Comp Replicated Amount of post-compressed data (in GiB). Network Bytes Amount of network throughput data (in GiB). Files Succeeded Number of files that were successfully replicated. Files Failed Number of files that failed to be replicated. Show in new window Brings up a separate window. Print Prints the graph.
Topology view The Topology view shows how the selected replication pairs are configured in the network. l The arrow – which is green (normal), yellow (warning), or red (error) – between DD systems represents one or more replication pairs. l To view details, select a context , which opens the Context Summary dialog, with links to Show Summary, Modify Options, Enable/Disable Pair, Graph Performance, and Delete Pair. l Select Collapse All to roll-up the Expand All context view and show only the name of the system and the count of destination contexts. l Select Expand All to show all the destination directory and MTree contexts configured on other systems.
Topology view 365 DD Replicator
l Select Reset Layout to return to the default view. l Select Print to open a standard print dialog.
Performance view The Performance view displays a graph that represents the fluctuation of data during replication. These are aggregated statistics of each replication pair for this DD system.
l Duration (x-axis) is 30 days by default. l Replication Performance (y-axis) is in GibiBytes or MebiBytes (the binary equivalents of GigaBytes and MegaBytes). l Network In is the total replication network bytes entering the system (all contexts). l Network Out is the total replication network bytes leaving the system (all contexts). l For a reading of a specific point in time, hover the cursor over a place on the graph. l During times of inactivity (when no data is being transferred), the shape of the graph may display a gradually descending line, instead of an expected sharply descending line.
Advanced Settings view Advanced Settings lets you manage throttle and network settings. Throttle Settings
l Temporary Override – Displays throttle rate if configured, or 0 meaning all replication traffic is stopped. l Permanent Schedule – Displays the time and days of the week on which scheduled throttling occurs.
l Bandwidth – Displays the configured data stream rate if bandwidth has been configured, or Unlimited (default) if not. The average data stream to the replication destination is at least 98,304 bits per second (12 KiB). l Delay – Displays the configured network delay setting (in milliseconds) if it has been configured, or None (default) if not. l Listen Port – Displays the configured listen port value if it has been configured, or 2051 (default) if not.
Adding throttle settings To modify the amount of bandwidth used by a network for replication, you can set a replication throttle for replication traffic. There are three types of replication throttle settings:
l Scheduled throttle – The throttle rate is set at a predetermined time or period. l Current throttle – The throttle rate is set until the next scheduled change, or until a system reboot. l Override throttle – The previous two types of throttle are overridden. This persists – even through reboot – until you select Clear Throttle Override or issue the replication throttle reset override command. You can also set a default throttle or a throttle for specific destinations, as follows:
366 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
l Default throttle – When configured, all replication contexts are limited to this throttle, except for those destinations specified by destination throttles (see next item). l Destination throttle – This throttle is used when only a few destinations need to be throttled, or when a destination requires a throttle setting different from the default throttle. When a default throttle already exists, this throttle takes precedence for the destination specified. For example, you can set the default replication throttle to 10 kbps, but – using a destination throttle – you can set a single collection replication context to unlimited.
Currently, you can set and modify destination throttle only by using the command- line interface (CLI); this functionality is not available in the DD System Manager. For documentation on this feature, see the replication throttle command in the EMC Data Domain Operating System Command Reference Guide. If the DD System Manager detects that you have one or more destination throttles set, you will be given a warning, and you should use the CLI to continue.
Additional notes about replication throttling:
l Throttles are set only at the source. The only throttle that applies to a destination is the 0 Bps (Disabled) option, which disables all replication traffic. l The minimum value for a replication throttle is 98,304 bits per second. Procedure
1. Select Replication > Advanced Settings > Add Throttle Setting to display the Add Throttle Setting dialog. 2. Set the days of the week for which throttling is to be active by selecting Every Day or by selecting checkbox(es) next to individual day(s). 3. Set the time that throttling is to start with the Start Time drop-down selectors for the hour:minute and AM/PM. 4. For Throttle Rate:
l Select Unlimited to set no limits. l Enter a number in the text box (for example, 20000), and select the rate from the menu (bps, Kbps, Bps, or KBps). l Select the 0 Bps (disabled) option to disable all replication traffic.
5. Select OK to set the schedule. The new schedule is shown under Permanent Schedule. Results Replication runs at the given rate until the next scheduled change, or until a new throttle setting forces a change.
Deleting Throttle Settings You can delete a single throttle setting or all throttle settings at once. Procedure
1. Select Replication > Advanced Settings > Delete Throttle Setting to display the Delete Throttle Setting dialog. 2. Select the checkbox for the throttle setting to delete, or select the heading checkbox to delete all settings. This list can include settings for the “disabled” state.
Advanced Settings view 367 DD Replicator
3. Select OK to remove the setting. 4. In the Delete Throttle Setting Status dialog, select Close.
Temporarily overriding a throttle setting A throttle override temporarily changes a throttle setting. The current setting is listed at the top of the window. Procedure
1. Select Replication > Advanced Settings > Set Throttle Override to display the Throttle Override dialog. 2. Either set a new throttle override, or clear a previous override. a. To set a new throttle override:
l Select Unlimited to revert to the system-set throttle rate (no throttling performed), or l Set the throttling bit and rate in the text box (for example, 20000) and (bps, Kbps, Bps, or KBps), or l Select 0 Bps (Disabled) to set the throttle rate to 0, effectively stopping all replication network traffic. l To enforce the change temporarily, select Clear at next scheduled throttle event.
b. To clear an override previously set, select Clear Throttle Override. 3. Select OK.
Changing network settings Using the bandwidth and network-delay settings together, replication calculates the proper TCP (transmission control protocol) buffer size for replication usage. These network settings are global to the DD system and should be set only once per system. Note the following:
l You can determine the actual bandwidth and the actual network delay values for each server by using the ping command.
l The default network parameters in a restorer work well for replication in low latency configurations, such as a local 100Mbps or 1000Mbps Ethernet network, where the latency round-trip time (as measured by the ping command) is usually less than 1 millisecond. The defaults also work well for replication over low- to moderate- bandwidth WANs, where the latency may be as high as 50-100 milliseconds. However, for high-bandwidth high-latency networks, some tuning of the network parameters is necessary. The key number for tuning is the bandwidth-delay number produced by multiplying the bandwidth and round-trip latency of the network. This number is a measure of how much data can be transmitted over the network before any acknowledgments can return from the far end. If the bandwidth-delay number of a replication network is more than 100,000, then replication performance benefits from setting the network parameters in both restorers. Procedure
1. Select Replication > Advanced Settings > Change Network Settings to display the Network Settings dialog. 2. In the Network Settings area, select Custom Values.
368 EMC Data Domain Operating System 5.7 Administration Guide DD Replicator
3. Enter Delay and Bandwidth values in the text boxes. The network delay setting is in milliseconds, and bandwidth is in bytes per second. 4. In the Listen Port area, enter a new value in the text box. The default IP Listen Port for a replication destination for receiving data streams from the replication source is 2051. This is a global setting for the DD system. 5. Select OK. The new settings appear in the Network Settings table.
Monitoring replication The DD System Manager provides many ways to track the status of replication – from checking replication pair status, to tracking backup jobs, to checking performance, to tracking a replication process.
Checking replication pair status Several places in the Replication area of the DD System Manager provide status updates for replication pairs. Procedure
1. Select Replication > Topology. 2. Check the colors of the arrows, which provide the status of the context. 3. Select the Summary tab. 4. From the Filter By drop-down list (under the Create Pair button), select State, and select Error, Warning, or Normal from the state menu. The replication contexts are sorted according to the selection.
Viewing estimated completion time for backup jobs You can use the Completion Predictor to see the estimated time for when a backup replication job will be completed. Procedure
1. Select Replication > Summary. 2. Select a Replication context for which to display Detailed Information. 3. In the Completion Predictor area, select options from the Source Time drop-down list for a replication’s completion time, and select Track. The estimated time displays, in the Completion Time area, for when a particular backup job will finish its replication to the destination. If the replication is finished, the area shows Completed.
Monitoring replication 369 DD Replicator
Checking replication context performance To check the performance of a replication context over time, select a Replication context in the Summary view, and select Performance Graph in the Detailed Information area.
Tracking status of a replication process To display the progress of a replication initialization, resynchronization, or recovery operation, use the Replication > Summary view to check the current state. CLI Equivalent # replication show config all CTX Source Destination Connection Host and Port Enabled ------1 dir://host2/backup/dir2 dir://host3/backup/dir3 host3.company.com Yes 2 dir://host3/backup/dir3 dir://host2/backup/dir2 host3.company.com Yes When specifying an IP version, use the following command to check its setting: # replication show config rctx://2
CTX: 2 Source: mtree://ddbeta1.dallasrdc.com/data/col1/EDM1 Destination: mtree://ddbeta2.dallasrdc.com/data/col1/EDM_ipv6 Connection Host: ddbeta2-ipv6.dallasrdc.com Connection Port: (default) Ipversion: ipv6 Low-bw-optim: disabled Encryption: disabled Enabled: yes Propagate-retention-lock: enabled
370 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 15 DD Secure Multitenancy
This chapter includes:
l Data Domain Secure Multitenancy overview. 372 l Provisioning a Tenant Unit. 375 l Enabling Tenant Self-Service mode. 378 l Data access by protocol. 379 l Data management operations. 381
DD Secure Multitenancy 371 DD Secure Multitenancy
Data Domain Secure Multitenancy overview Data Domain Secure Multitenancy (SMT) is the simultaneous hosting, by an internal IT department or an external provider, of an IT infrastructure for more than one consumer or workload (business unit, department, or Tenant). SMT provides the ability to securely isolate many users and workloads in a shared infrastructure, so that the activities of one Tenant are not apparent or visible to the other Tenants. A Tenant is a consumer (business unit, department, or customer) who maintains a persistent presence in a hosted environment. Within an enterprise, a Tenant may consist of one or more business units or departments on a DD system configured and managed by IT staff.
l For a business unit (BU) use case, the Finance and Human Resources departments of a corporation could share the same DD system, but each department would be unaware of the presence of the other. l For a service provider (SP) use case, the SP could deploy one or more DD systems to accommodate different Protection Storage services for multiple end-customers. Both use cases emphasize the segregation of different customer data on the same physical DD system.
SMT architecture basics Secure Multitenancy (SMT) provides a simple approach to setting up Tenants and Tenant Units, using MTrees. SMT setup is performed using DD Management Center (DDMC) and/or the DD OS command line interface. This administration guide provides the theory of SMT and some general command line instructions. The basic architecture of SMT is as follows.
l A Tenant is created on the DD Management Center and/or DD system. l A Tenant Unit is created on a DD system for the Tenant. l One or more MTrees are created to meet the storage requirements for the Tenant's various types of backups. l The newly created MTrees are added to the Tenant Unit. l Backup applications are configured to send each backup to its configured Tenant Unit MTree.
For more information about DD Management Center, see the DD Management Center User Guide. For more information about the DD OS command line interface, see the DD OS Command Reference.
Terminology used in SMT Understanding the terminology used in SMT (Secure Multitenancy) will help you better understand this unique environment. MTrees MTrees are logical partitions of the Data Domain file system and offer the highest degree of management granularity, meaning users can perform operations on a specific MTree
372 EMC Data Domain Operating System 5.7 Administration Guide DD Secure Multitenancy
without affecting the entire file system. MTrees are assigned to Tenant Units and contain that Tenant Unit's individualized settings for managing and monitoring SMT. Multitenancy Multitenancy refers to the hosting of an IT infrastructure by an internal IT department, or an external service provider, for more than one consumer/workload (business unit/ department/Tenant) simultaneously. Data Domain SMT enables Data Protection-as-a- Service. RBAC (role-based access control) RBAC (role-based access control offers multiple roles with different privilege levels, which combine to provide the administrative isolation on a multitenant DD system. (The next section will define these roles.) Storage Unit A Storage Unit is an MTree configured for the DD Boost protocol. Data isolation is achieved by creating a Storage Unit and assigning it to a DD Boost user. The DD Boost protocol permits access only to Storage Units assigned to DD Boost users connected to the DD system. Tenant A Tenant is a consumer (business unit/department/customer) who maintains a persistent presence in a hosted environment. Tenant Self-Service Tenant Self-Service is a method of letting a Tenant log in to a DD system to perform some basic services (add, edit, or delete local users, NIS groups, and/or AD groups). This reduces the bottleneck of always having to go through an administrator for these basic tasks. The Tenant can access only their assigned Tenant Units. Tenant Users and Tenant Admins will, of course, have different privileges. Tenant Unit A Tenant Unit is the partition of a DD system that serves as the unit of administrative isolation between Tenants. Tenant Units are secured and logically isolated from each other, which ensures security and isolation of the control path when running multiple Tenants simultaneously on the shared infrastructure. Tenant Units can contain one or more MTrees, which hold all configuration elements needed in a multitenancy setup. Users, management-groups, notification-groups, and other configuration elements are part of a Tenant Unit.
Control path and administrative isolation Control path isolation is achieved by providing the user roles of tenant-admin and tenant- user for a Tenant Unit. Administrative isolation is achieved by associating a fixed set of management IP address(es) with a Tenant Unit. The tenant-admin and tenant-user roles are restricted in scope and capability to specific Tenant Units and to a restricted set of operations they can perform on those Tenant Units. To ensure a logically secure and isolated data path, a system administrator must configure one or more Tenant Unit MTrees for each protocol in an SMT environment. Supported protocols include DD Boost, NFS, CIFS, and VTL. Access is strictly regulated by the native access control mechanisms of each protocol. Tenant-self-service sessions (through ssh) can be restricted to a fixed set of management IP address(es) on a DD system. Administrative access sessions (through ssh/http/https) can also be restricted to a fixed set of management IP address(es) on DD systems. By default, however, there are no management IP address(es) associated with a Tenant Unit, so the only standard restriction is through the use of the tenant-admin and tenant-user
Control path and administrative isolation 373 DD Secure Multitenancy
roles. You must use smt tenant-unit management-ip to add and maintain management IP address(es) for Tenant Units.
Understanding RBAC in SMT In SMT (Secure Multitenancy), permission to perform a task depends on the role assigned to a user. DD Management Center uses RBAC (role-based access control) to control these permissions. All DD Management Center users can: l View all Tenants l Create, read, update, or delete Tenant Units belonging to any Tenant if the user is an administrator on the DD system hosting the Tenant Unit l Assign and unassign Tenant Units to and from a Tenant if the user is an administrator on the DD system hosting the Tenant Unit l View Tenant Units belonging to any Tenant if the user has any assigned role on the DD system hosting the Tenant Unit To perform more advanced tasks depends on the role of the user, as follows: admin role A user with an admin role can perform all administrative operations on a DD system. An admin can also perform all SMT administrative operations on a DD system, including setting up SMT, assigning SMT user roles, enabling Tenant self-service mode, creating a Tenant, and so on. In the context of SMT, the admin is typically referred to as the landlord. In DD OS, the role is known as the sysadmin. To have permission to edit or delete a Tenant, you must be both a DD Management Center admin and a DD OS sysadmin on all DD systems associated with the Tenant Units of that Tenant. If the Tenant does not have any Tenant Units, you need only to be a DD Management Center admin to edit or delete that Tenant. tenant-admin role A user with a tenant-admin role can perform certain tasks only when tenant self-service mode is enabled for a specific Tenant Unit. Responsibilities include scheduling and running a backup application for the Tenant and monitoring resources and statistics within the assigned Tenant Unit. In addition, tenant-admins ensure administrative separation when Tenant self-service mode is enabled. In the context of SMT, the tenant- admin is usually referred to as the backup admin. tenant-user role A user with a tenant-user role can monitor the performance and usage of SMT components only on Tenant Unit(s) assigned to them and only when Tenant self-service is enabled. In addition, tenant-users may run the show and list commands. none role A user with a role of none is not allowed to perform any operations on a DD system other than changing their password and accessing data using DD Boost. However, after SMT is enabled, the admin can select a user with a none role from the DD system and assign them an SMT-specific role of tenant-admin or tenant-user. Then, that user can perform operations on SMT management objects. management groups BSPs (backup service providers) can use management groups defined in a single, external AD (active directory) or NIS (network information service) to simplify managing user roles on Tenant Units. Each BSP Tenant may be a separate, external company and may use a name-service such as AD or NIS. With SMT management groups, the AD and NIS servers are set up and configured by the admin in the same way as SMT local users. The admin can ask their AD or NIS
374 EMC Data Domain Operating System 5.7 Administration Guide DD Secure Multitenancy
administrator to create and populate the group. The admin then assigns an SMT role to the entire group. Any user within the group who logs in to the DD system is logged in with the role assigned to the group. When users leave or join a Tenant company, they can be removed or added to the group by the AD or NIS administrator. It is not neceesary to modify the RBAC configuration on a DD system when users who are part of the group are added or removed.
Provisioning a Tenant Unit Launching the configuration wizard begins the initial provisioning procedure for Secure Multitenancy (SMT). During the procedure, the wizard creates and provisions a new Tenant Unit based on Tenant configuration requirements. Information is entered by the administrator, as prompted. After completing the procedure, the administrator proceeds to the next set of tasks, beginning with enabling Tenant Self-Service mode. Following the initial setup, manual procedures and configuration modifications may be performed as required. Procedure 1. Start SMT. # smt enable SMT enabled. 2. Verify that SMT is enabled. # smt status SMT is enabled. 3. Launch the SMT configuration wizard. # smt tenant-unit setup No tenant-units. 4. Follow the configuration prompts. SMT TENANT-UNIT Configuration
Configure SMT TENANT-UNIT at this time (yes|no) [no]: yes
Do you want to create new tenant-unit (yes/no)? : yes
Tenant-unit Name Enter tenant-unit name to be created : SMT_5.7_tenant_unit Invalid tenant-unit name. Enter tenant-unit name to be created : SMT_57_tenant_unit
Pending Tenant-unit Settings Create Tenant-unit SMT_57_tenant_unit
Do you want to save these settings (Save|Cancel|Retry): save SMT Tenant-unit Name Configurations saved.
SMT TENANT-UNIT MANAGEMENT-IP Configuration
Configure SMT TENANT-UNIT MANAGEMENT-IP at this time (yes|no) [no]: yes
Do you want to add a local management ip to this tenant-unit? (yes|no) [no]: yes
port enabled state DHCP IP address netmask type additional /prefix length setting ------ethMa yes running no 192.168.10.57 255.255.255.0 n/a fe80::260:16ff:fe49:f4b0** /64 eth3a yes running ipv4 192.168.10.236* 255.255.255.0* n/a fe80::260:48ff:fe1c:60fc** /64
Provisioning a Tenant Unit 375 DD Secure Multitenancy
eth3b yes running no 192.168.50.57 255.255.255.0 n/a fe80::260:48ff:fe1c:60fd** /64 eth4b yes running no 192.168.60.57 255.255.255.0 n/a fe80::260:48ff:fe1f:5183** /64 ------* Value from DHCP ** auto_generated IPv6 address
Choose an ip from above table or enter a new ip address. New ip addresses will need to be created manually.
Ip Address Enter the local management ip address to be added to this tenant-unit : 192.168.10.57
Do you want to add a remote management ip to this tenant-unit? (yes|no) [no]:
Pending Management-ip Settings
Add Local Management-ip 192.168.10.57 Do you want to save these settings (Save|Cancel|Retry): yes unrecognized input, expecting one of Save|Cancel|Retry
Do you want to save these settings (Save|Cancel|Retry): save Local management access ip "192.168.10.57" added to tenant-unit "SMT_57_tenant_unit".
SMT Tenant-unit Management-IP Configurations saved.
SMT TENANT-UNIT MANAGEMENT-IP Configuration
Do you want to add another local management ip to this tenant-unit? (yes|no) [no]:
Do you want to add another remote management ip to this tenant-unit? (yes|no) [no]:
SMT TENANT-UNIT DDBOOST Configuration Configure SMT TENANT-UNIT DDBOOST at this time (yes|no) [no]:
SMT TENANT-UNIT MTREE Configuration Configure SMT TENANT-UNIT MTREE at this time (yes|no) [no]: yes
Name Pre-Comp (GiB) Status Tenant-Unit ------/data/col1/laptop_backup 4846.2 RO/RD - /data/col1/random 23469.9 RO/RD - /data/col1/software2 2003.7 RO/RD - /data/col1/tsm6 763704.9 RO/RD ------D : Deleted Q : Quota Defined RO : Read Only RW : Read Write RD : Replication Destination RLGE : Retention-Lock Governance Enabled RLGD : Retention-Lock Governance Disabled RLCE : Retention-Lock Compliance Enabled
Do you want to assign an existing MTree to this tenant-unit? (yes|no) [no]:
Do you want to create a mtree for this tenant-unit now? (yes|no) [no]: yes
MTree Name Enter MTree name : SMT_57_tenant_unit Invalid mtree path name. Enter MTree name : SMT_57_tenant_unit
Invalid mtree path name.
376 EMC Data Domain Operating System 5.7 Administration Guide DD Secure Multitenancy
Enter MTree name : /data/col1/SMT_57_tenant_unit
MTree Soft-Quota Enter the quota soft-limit to be set on this MTree (
MTree Hard-Quota Enter the quota hard-limit to be set on this MTree (
Pending MTree Settings Create MTree /data/col1/SMT_57_tenant_unit MTree Soft Limit none MTree Hard Limit none
Do you want to save these settings (Save|Cancel|Retry): save MTree "/data/col1/SMT_57_tenant_unit" created successfully. MTree "/data/col1/SMT_57_tenant_unit" assigned to tenant-unit "SMT_57_tenant_unit".
SMT Tenant-unit MTree Configurations saved.
SMT TENANT-UNIT MTREE Configuration
Name Pre-Comp (GiB) Status Tenant-Unit ------/data/col1/laptop_backup 4846.2 RO/RD - /data/col1/random 23469.9 RO/RD - /data/col1/software2 2003.7 RO/RD - /data/col1/tsm6 763704.9 RO/RD ------D : Deleted Q : Quota Defined RO : Read Only RW : Read Write RD : Replication Destination RLGE : Retention-Lock Governance Enabled RLGD : Retention-Lock Governance Disabled RLCE : Retention-Lock Compliance Enabled
Do you want to assign another MTree to this tenant-unit? (yes|no) [no]: yes
Do you want to assign an existing MTree to this tenant-unit? (yes|no) [no]:
Do you want to create another mtree for this tenant-unit? (yes|no) [no]:
SMT TENANT-UNIT SELF-SERVICE Configuration
Configure SMT TENANT-UNIT SELF-SERVICE at this time (yes|no) [no]: yes Self-service of this tenant-unit is disabled
Do you want to enable self-service of this tenant-unit? (yes|no) [no]: yes
Do you want to configure a management user for this tenant-unit? (yes|no) [no]:
Do you want to configure a management group for this tenant-unit (yes|no) [no]: yes
Management-Group Name Enter the group name to be assigned to this tenant-unit : SMT_57_tenant_unit_group
What role do you want to assign to this group (tenant-user|tenant-admin) [tenant-user]: tenant-admin
Management-Group Type What type do you want to assign to this group (nis|active-directory)? : nis
Pending Self-Service Settings Enable Self-Service SMT_57_tenant_unit
Provisioning a Tenant Unit 377 DD Secure Multitenancy
Assign Management-group SMT_57_tenant_unit_group Management-group role tenant-admin Management-group type nis
Do you want to save these settings (Save|Cancel|Retry): save Tenant self-service enabled for tenant-unit "SMT_57_tenant_unit" Management group "SMT_57_tenant_unit_group" with type "nis" is assigned to tenant-unit "SMT_57_tenant_unit" as "tenant-admin".
SMT Tenant-unit Self-Service Configurations saved.
SMT TENANT-UNIT SELF-SERVICE Configuration
Do you want to configure another management user for this tenant-unit? (yes|no) [no]:
Do you want to configure another management group for this tenant-unit? (yes|no) [no]:
SMT TENANT-UNIT ALERT Configuration
Configure SMT TENANT-UNIT ALERT at this time (yes|no) [no]: yes No notification lists.
Alert Group Name Specify alert notify-list group name to be created : SMT_57_tenant_unit_notify
Alert email addresses Enter email address to receive alert for this tenant-unit : [email protected]
Do you want to add more emails (yes/no)? : no
Pending Alert Settings Create Notify-list group SMT_57_tenant_unit_notify Add emails [email protected]
Do you want to save these settings (Save|Cancel|Retry): save Created notification list "SMT_57_tenant_unit_notify" for tenant "SMT_57_tenant_unit". Added emails to notification list "SMT_57_tenant_unit_notify".
SMT Tenant-unit Alert Configurations saved.
Enabling Tenant Self-Service mode For administrative separation of duties and delegation of administrative/management tasks to implement Tenant Self-Service, which is required for control path isolation, the system administrator can enable this mode on a Tenant Unit and then assign users to manage the unit in the roles of tenant-admin or tenant-user. These roles allow users other than the administrator to perform specific tasks on the Tenant Unit to which they are assigned. In addition to administrative separation, Tenant Self-Service mode helps reduce the management burden on internal IT and service provider staff. Procedure 1. View Tenant Self-Service mode status for one or all Tenant Units. # smt tenant-unit option show < tenant-unit | all >2. Enable Tenant Self-Service mode on the selected Tenant Unit. # smt tenant-unit option set tenant-unit self-service
378 EMC Data Domain Operating System 5.7 Administration Guide DD Secure Multitenancy
Data access by protocol Secure data paths, with protocol-specific access controls, enable security and isolation for Tenant Units. In a Secure Multitenancy (SMT) environment, data access protocol management commands are also enhanced with a Tenant Unit parameter to enable consolidated reporting. DD systems support multiple data access protocols simultaneously, including DD Boost, NFS, CIFS, and VTL. A DD system can present itself as an application-specific interface, such as a file server offering NFS or CIFS access over the Ethernet, a VTL device, or a DD Boost device. The native access control mechanisms of each supported protocol ensure that the data paths for each Tenant remain separate and isolated. Such mechanisms include access control lists (ACLs) for CIFS, exports for NFS, DD Boost credentials, and Multi-User Boost credential-aware access control.
Multi-User DD Boost and Storage Units in SMT When using Multi-User DD Boost with SMT (Secure Multitenancy), user permissions are set by Storage Unit ownership. Multi-User DD Boost refers to the use of multiple DD Boost user credentials for DD Boost Access Control, in which each user has a separate user name and password. A Storage Unit is an MTree configured for the DD Boost protocol. A user can be associated with, or “own,” one or more Storage Units. Storage Units that are owned by one user cannot be owned by another user. Therefore, only the user owning the Storage Unit can access the Storage Unit for any type of data access, such as backup/restore. The number of DD Boost user names cannot exceed the maximum number of MTrees. (See the “MTrees” chapter in this book for the current maximum number of MTrees for each DD model.) Storage Units associated with SMT must have the none role assigned to them. Each backup application must authenticate using its DD Boost user name and password. After authentication, DD Boost verifies the authenticated credentials to confirm ownership of the Storage Unit. The backup application is granted access to the Storage Unit only if the user credentials presented by the backup application match the user names associated with the Storage Unit. If user credentials and user names do not match, the job fails with a permission error.
Configuring access for CIFS Common Internet File System (CIFS) is a file-sharing protocol for remote file access. In a Secure Multitenancy (SMT) configuration, backup and restores require client access to the CIFS shares residing in the MTree of the associated Tenant Unit. Data isolation is achieved using CIFS shares and CIFS ACLs. Procedure 1. Create an MTree for CIFS. # mtree create mtree-path tenant-unit tenant-unit] 2. Assign the MTree to the Tenant Unit. # mtree modify mtree-path tenant-unit tenant-unit | none] 3. Set capacity soft and hard quotas for the MTree. # mtree create mtree-path tenant-unit tenant-unit] [quota-soft- limit n
Data access by protocol 379 DD Secure Multitenancy
4. Create a CIFS share for pathname from the MTree. # cifs share create share path pathname clients clients
Configuring NFS access NFS is a UNIX-based, file-sharing protocol for remote file access. In a Secure Multitenancy (SMT) environment, backup and restores require client access to the NFS exports residing in the MTree of the associated Tenant Unit. Data isolation is achieved using NFS exports and network isolation. NFS determines if an MTree is associated with a network-isolated Tenant Unit. If so, NFS verifies the connection properties associated with the Tenant Unit. Connection properties include the destination IP address and interface or client hostname. Procedure 1. Create an MTree for NFS. # mtree create mtree-path tenant-unit tenant-unit] 2. Assign the MTree to the Tenant Unit. # mtree modify mtree-path tenant-unit tenant-unit | none] 3. Set capacity soft and hard quotas for the MTree. # mtree create mtree-path tenant-unit tenant-unit] [quota-soft- limit n
Configuring access for VTL VTL Tenant data isolation is achieved using VTL access groups that create a virtual access path between a host system and the VTL. (The physical Fibre Channel connection between the host system and VTL must already exist.) Placing tapes in the VTL allows them to be written to, and read by, the backup application on the host system. VTL tapes are created in a VTL pool, which is an MTree. Because VTL pools are MTrees, the pools can be assigned to Tenant Units. This association enables SMT monitoring and reporting. For example, if a tenant-admin is assigned a Tenant Unit that contains a VTL pool, the tenant-admin can run MTree commands to display read-only information. Commands can run only on the VTL pool assigned to the Tenant Unit. These commands include: l mtree list to view a list of MTrees in the Tenant Unit l mtree show compression to view statistics on MTree compression l mtree show performance to view statistics on performance Output from most list and show commands include statistics that enable service providers to measure space usage and calculate chargeback fees. VTL operations are unaffected and continue to function normally.
380 EMC Data Domain Operating System 5.7 Administration Guide DD Secure Multitenancy
Using VTL NDMP TapeServer VTL Tenant data isolation is also achieved using NDMP. DD OS implements a NDMP (Network Data Management Protocol) tape server that allows NDMP-capable systems to send backup data to the DD system via a three-way NDMP backup. The backup data is written to virtual tapes (which are in a pool) by a VTL assigned to the special VTL group TapeServer. Because the backup data is written to tapes in a pool, information in the VTL topic regarding MTrees also applies to the Data Domain NDMP TapeServer.
Data management operations Secure Multitenancy (SMT) management operations include monitoring Tenant Units and other objects, such as Storage Units and MTrees. For some SMT objects, additional configuration or modification may also be required.
Collecting performance statistics Each MTree can be measured for performance or “usage” statistics and other real-time information. Historical consumption rates are available for DD Boost Storage Units. Command output lets the tenant-admin collect usage statistics and compression ratios for an MTree associated with a Tenant Unit, or for all MTrees and associated Tenant Units. Output may be filtered to display usage in intervals ranging from minutes to months. Results are passed to the administrator, who uses the statistics as a chargeback metric. A similar method is used to gather usage statistics and compression ratios for Storage Units. Procedure 1. Collect MTree real-time performance statistics. # mtree show stats 2. Collect performance statistics for MTrees associated with a Tenant Unit. # mtree show performance 3. Collect compression statistics for MTrees associated with a Tenant Unit. # mtree show compression
Modifying quotas To meet QoS criteria, a system administrator uses DD OS “knobs” to adjust the settings required by the Tenant configuration. For example, the administrator can set “soft” and “hard” quota limits on DD Boost Storage Units. Stream “soft” and “hard” quota limits can be allocated only to DD Boost Storage Units assigned to Tenant Units. After the administrator sets the quotas, the tenant-admin can monitor one or all Tenant Units to ensure no single object exceeds its allocated quotas and deprives others of system resources. Quotas are set initially when prompted by the configuration wizard, but they can be adjusted or modified later. The example below shows how to modify quotas for DD Boost. (You can also use quota capacity and quota streams to deal with capacity and stream quotas and limits.) Procedure 1. To modify soft and hard quota limits on DD Boost Storage Unit “su33”:
Using VTL NDMP TapeServer 381 DD Secure Multitenancy
ddboost storage-unit modify su33 quota-soft-limit 10 Gib quota- hard-limit 20 Gib 2. To modify stream soft and hard limits on DD Boost Storage Unit “su33”:
ddboost storage-unit modify su33 write-stream-soft-limit 20 read- stream-soft-limit 6 repl -stream-soft-limit 20 combined-stream- soft-limit 20 3. To report physical size for DD Boost Storage Unit “su33”:
ddboost storage-unit modify su33 report-physical-size 8 GiB
SMT and replication In case of disaster, user roles dictate how a user can assist in data recovery operations. Several replication types are available in an SMT configuration. (See the DD Replicator chapter for more detail on how to perform replication.) Here are some points to consider regarding user roles:
l The admin can recover MTrees from a replicated copy. l The tenant-admin can replicate MTrees from one system to another, using DD Boost managed file replication. l The tenant-admin can recover MTrees from a replicated copy, also by using DD Boost managed file replication. Collection replication Collection replication replicates core Tenant Unit configuration information. MTree replication (NFS/CIFS) using DD Boost managed file replication MTree replication is supported on MTrees assigned to Tenant Units, using DD Boost managed file replication. During MTree replication, an MTree assigned to a Tenant Unit on one system can be replicated to an MTree assigned to a Tenant Unit on another system. MTree replication is not allowed between two different Tenants on the two DD systems. When security mode is set to strict, MTree replication is allowed only when the MTrees belong to same Tenants. For backward compatibility, MTree replication from an MTree assigned to a Tenant Unit to an unassigned MTree is supported, but must be configured manually. Manual configuration ensures the destination MTree has the correct settings for the Tenant Unit. Conversely, MTree replication from an unassigned MTree to an MTree assigned to a Tenant Unit is also supported. When setting up SMT-aware MTree replication, security mode defines how much checking is done on the Tenant. The default mode checks that the source and destination do not belong to different Tenants. The strict mode makes sure the source and destination belong to the same Tenant. Therefore, when you use strict mode, you must create a Tenant on the destination machine with the same UUID as the UUID of the Tenant on the source machine that is associated with the MTree being replicated. DD Boost managed file replication (also with DD Boost AIR) DD Boost managed file replication is supported between Storage Units, regardless of whether one Storage Unit, or both, are assigned to Tenant Units. During DD Boost managed file replication, Storage Units are not replicated in total. Instead, certain files within a Storage Unit are selected by the backup application for replication. The files selected in a Storage Unit and assigned to a Tenant Unit on one system can be replicated to a Storage Unit assigned to a Tenant Unit on another system.
382 EMC Data Domain Operating System 5.7 Administration Guide DD Secure Multitenancy
For backward compatibility, selected files in a Storage Unit assigned to a Tenant Unit can be replicated to an unassigned Storage Unit. Conversely, selected files in an unassigned Storage Unit can be replicated to a Storage Unit assigned to a Tenant Unit. DD Boost managed file replication can also be used in DD Boost AIR deployments.
SMT Tenant alerts A DD system generates events when it encounters potential problems with software or hardware. When an event is generated, an alert notification is sent immediately via email to members designated in the notification list and to the Data Domain administrator. SMT alerts are specific to each Tenant Unit and differ from DD system alerts. When Tenant Self-Service mode is enabled, the tenant-admin can choose to receive alerts about the various system objects he or she is associated with and any critical events, such as an unexpected system shutdown. A tenant-admin may only view or modify notification lists to which he or she is associated. The example below shows a sample alert. Notice that the two event messages at the bottom of the notification are specific to a Multi-Tenant environment (indicated by the word “Tenant”). For the entire list of DD OS and SMT alerts, see the EMC Data Domain MIB Quick Reference Guide or the SNMP MIB. EVT-ENVIRONMENT-00021 – Description: The system has been shutdown by abnormal method; for example, not by one of the following: 1) Via IPMI chassis control command 2) Via power button 3) Via OS shutdown.
Action: This alert is expected after loss of AC (main power) event. If this shutdown is not expected and persists, contact your contracted support provider or visit us online at https:// my.datadomain.com.
Tenant description: The system has experienced an unexpected power loss and has restarted.
Tenant action: This alert is generated when the system restarts after a power loss. If this alert repeats, contact your System Administrator.
Managing snapshots A snapshot is a read-only copy of an MTree captured at a specific point in time. A snapshot can be used for many things, for example, as a restore point in case of a system malfunction. The required role for using snapshot is admin or tenant-admin. To view snapshot information for an MTree or a Tenant Unit: # snapshot list mtree mtree-path | tenant-unit tenant-unit To view a snapshot schedule for an MTree or a Tenant Unit: # snapshot schedule show [name | mtrees mtree-listmtree-list | tenant- unit tenant-unit]
Performing a file system Fast Copy A Fast Copy operation clones files and directory trees of a source directory to a target directory on a DD system. There are special circumstances regarding Fast Copy with Secure Multitenancy (SMT). Here are some considerations when performing a file system Fast Copy with Tenant Self- Service mode enabled:
SMT Tenant alerts 383 DD Secure Multitenancy
l A tenant-admin can Fast Copy files from one Tenant Unit to another when the tenant- admin is the tenant-admin for both Tenant Units, and the two Tenant Units belong to the same Tenant. l A tenant-admin can Fast Copy files within the same Tenant Unit. l A tenant-admin can Fast Copy files within the Tenant Units at source and destination. To perform a file system Fast Copy: # filesys fastcopy source destination
384 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 16 DD Extended Retention
This chapter includes:
l DD Extended Retention overview. 386 l Supported protocols in DD Extended Retention. 388 l High Availability and Extended Retention. 388 l Using DD Replicator with DD Extended Retention. 388 l Hardware and licensing for DD Extended Retention. 390 l Managing DD Extended Retention. 393 l Upgrades and recovery with DD Extended Retention. 403
DD Extended Retention 385 DD Extended Retention
DD Extended Retention overview EMC Data Domain Extended Retention (DD Extended Retention) provides an internal tiering approach that enables cost-effective, long-term retention of backup data on a DD system. DD Extended Retention lets you leverage DD systems for long-term backup retention and minimize reliance on tape.
DD Extended Retention was formerly known as Data Domain Archiver.
Two-Tiered File System The internal two-tiered file system of a DD Extended Retention-enabled DD system consists of an active tier and a retention tier. The file system, however, appears as a single entity. Incoming data is first placed in the active tier of the file system. The data (in the form of complete files) is later moved to the retention tier of the file system, as specified by your individual Data Movement Policy. For example, the active tier might retain weekly full and daily incremental backups for 90 days, while the retention tier might retain monthly fulls for seven years. The retention tier is comprised of one or more retention units, each of which may draw storage from one or more shelves.
As of DD OS 5.5.1, only one retention unit per retention tier is allowed. However, systems set up prior to DD OS 5.5.1 may continue to have more than one retention unit, but you will not be allowed to add any more retention units to them.
Transparency of Operation DD Extended Retention-enabled DD systems support existing backup applications using simultaneous data access methods through NFS and CIFS file service protocols over Ethernet, through VTL for open systems and IBMi, or as a disk-based target using application-specific interfaces, such as DD Boost (for use with EMC Avamar®, EMC NetWorker®, EMC GreenPlum, Symantec OpenStorage, and Oracle RMAN). DD Extended Retention extends the DD architecture with automatic transparent data movement from the active tier to the retention tier. All of the data in the two tiers is accessible, although there might be a slight delay on initial access to data in the retention tier. The namespace of the system is global and is not affected by data movement. No partitioning of the file system is necessary to take advantage of the two- tiered file system. Data Movement Policy The Data Movement Policy, which you can customize, is the policy by which files are moved from the active to the retention tier. It is based on the time when the file was last modified. You can set a different policy for each different subset of data, because the policy can be set on a per-MTree basis. Files that may be updated need a policy different from those that never change. Deduplication within Retention Unit For fault isolation purposes, deduplication occurs entirely within the retention unit for DD Extended Retention-enabled DD systems. There is no cross-deduplication between active and retention tiers, or between different retention units (if applicable).
386 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
Storage Drawn from Each Tier The concept of tiering extends to the storage level for a DD Extended Retention-enabled DD system. The active tier of the file system draws storage from the active tier of storage. The retention tier of the file system draws storage from the retention tier of storage.
For both active and retention tiers, DD OS 5.2 and later releases support ES20 and ES30 shelves. Different Data Domain controller models cannot be mixed in the same shelf set, and the shelf sets must be balanced according to the configuration rules specified in the EMC ES30 Expansion Shelf Hardware Guide. DD OS 5.7 and later support, DS60 shelves and the shelf sets must be balanced according to the configuration rules specified in the EMC DS60 Expansion Shelf Hardware Guide. You can attach significantly more storage to the same controller, for example, you can attach up to a maximum of 56 shelves on a DD990, when you are using DD Extended Retention. The active tier must include storage consisting of at least one shelf. For the minimum and maximum shelf configuration for the Data Domain controller models, refer to the expansion shelf hardware guides for ES30 and DS60.
Data Protection On a DD Extended Retention-enabled DD system, data is protected with built-in fault isolation features, disaster recovery capability, and DIA (Data Invulnerability Architecture). DIA checks files when they are moved from the active to the retention tier. After data is copied into the retention tier, the container and file system structures are read back and verified. The location of the file is updated, and the space on the active tier is reclaimed after the file is verified to have been correctly written to the retention tier. When a retention unit is filled up, namespace information and system files are copied into it, so the data in the retention unit may be recovered even when other parts of the system are lost.
Sanitization and some forms of Replication are not supported for DD Extended Retention- enabled DD systems.
Space Reclamation To reclaim space that has been freed up by data moved to the retention tier, you can use Space Reclamation (as of DD OS 5.3), which runs in the background as a low-priority activity. It suspends itself when there are higher priority activities, such as data movement and cleaning. Encryption of Data at Rest As of DD OS 5.5.1, you can use the Encryption of Data at Rest feature on DD Extended Retention-enabled DD systems, if you have an encryption license. Encryption is not enabled by default. This is an extension of the encryption capability already available, prior to DD OS 5.5.1, for systems not using DD Extended Retention. Refer to the Managing Encryption of Data at Rest chapter in this guide for complete instructions on setting up and using the encryption feature.
DD Extended Retention overview 387 DD Extended Retention
Supported protocols in DD Extended Retention DD Extended Retention-enabled DD systems support the protocols NFS, CIFS, and DD Boost. Support for VTL was added in DD OS 5.2, and support for NDMP was added in DD OS 5.3.
For a list of applications supported with DD Boost, see the DD Boost Compatibility List on the EMC Online Support site.
When you are using DD Extended Retention, data first lands in the active tier. Files are moved in their entirety into the retention unit in the retention tier, as specified by your Data Movement Policy. All files appear in the same namespace. There is no need to partition data, and you can continue to expand the file system as desired. All data is visible to all users, and all file system metadata is present in the active tier. The trade-off in moving data from the active to the retention tier is larger capacity versus slightly slower access time if the unit to be accessed is not currently ready for access.
High Availability and Extended Retention Data Domain systems with High Availability (HA) enabled do not support DD Extended Retention. DD OS 5.7.1 cannot currently support the Extended Retention Tier.
Using DD Replicator with DD Extended Retention Some forms of replication are supported on DD Extended Retention-enabled DD systems. Supported replication types depend on the data to be protected: l To protect data on a system as a source, a DD Extended Retention-enabled DD system supports collection replication, MTree replication, and DD Boost managed file replication. l To protect data from other systems as a destination, a DD Extended Retention- enabled DD system also supports directory replication, as well as collection replication, MTree replication, and DD Boost managed file replication.
Delta (low bandwidth optimization) replication is not supported with DD Extended Retention. You must disable delta replication on all contexts before enabling DD Extended Retention on a DD system.
Collection replication with DD Extended Retention Collection replication takes place between the corresponding active tier and retention unit of the two DD systems with DD Extended Retention enabled. If the active tier or retention unit at the source fails, the data can be copied from the corresponding unit at the remote site onto a new unit, which is shipped to your site as a replacement unit. Prerequisites for setting up collection replication include: l Both the source and destination systems must be configured as DD systems with DD Extended Retention enabled.
388 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
l The file system must not be enabled on the destination until the retention unit has been added to it, and replication has been configured.
Directory replication with DD Extended Retention For directory replication, a DD Extended Retention-enabled DD system serves as a replication target and supports one-to-one and many-to-one topologies from any supported DD system. However, DD Extended Retention-enabled DD systems do not support bi-directional directory replication and cannot be a source of directory replication.
To copy data using directory replication into a DD Extended Retention-enabled DD system, the source must be running DD OS 5.0 or later. Therefore, on systems running DD OS 5.0 or earlier, you must first import data into an intermediate system running DD OS 5.0 or later. For example, replication from a DD OS 4.9 Extended Retention-enabled system could be made into a DD OS 5.2 non-Extended Retention-enabled system. Then, replication could be made from the DD OS 5.2 system into the DD OS 4.9 system.
MTree replication with DD Extended Retention You can set up MTree replication between two DD Extended Retention-enabled DD systems. Replicated data is first placed in the active tier on the destination system. The Data Movement Policy on the destination system then determines when the replicated data is moved to the retention tier. Note that MTree replication restrictions and policies vary by DD OS release, as follows:
l As of DD OS 5.1, data can be replicated from a non-DD Extended Retention-enabled system to a DD Extended Retention-enabled system with MTree replication. l As of DD OS 5.2, data can be protected within an active tier by replicating it to the active tier of a DD Extended Retention-enabled system. l As of DD OS 5.5, MTree replication is supported from a DD Extended Retention- enabled system to a non-DD Extended Retention-enabled system if both are running DD OS 5.5 or later. l For DD OS 5.3 and 5.4, if you plan to enable DD Extended Retention, do not set up replication for the /backup MTree on the source machine. (DD OS 5.5 and later do not have this restriction.)
Managed file replication with DD Extended Retention For DD Extended Retention-enabled DD systems, the supported topologies for DD Boost managed file replication are one-to-one, many-to-one, bi-directional, one-to-many, and cascaded.
For DD Boost 2.3 or later, you can specify how multiple copies are to be made and managed within the backup application.
Directory replication with DD Extended Retention 389 DD Extended Retention
Hardware and licensing for DD Extended Retention Certain hardware configurations are required for DD Extended Retention-enabled DD systems. Licensing, specifically separate shelf capacity licenses, is also specific to this feature.
Hardware supported for DD Extended Retention The hardware requirements for DD Extended Retention-enabled DD systems include memory requirements, shelves, NIC/FC cards, and so on. For details about the required hardware configurations for DD Extended Retention, see the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. The following DD systems support DD Extended Retention: DD860
l 72 GB of RAM l 1 - NVRAM IO module (1 GB) l 3 - Quad-port SAS IO modules l 2 - 1 GbE ports on the motherboard l 0 to 2 - 1/10 GbE NIC IO cards for external connectivity l 0 to 2 - Dual-Port FC HBA IO cards for external connectivity l 0 to 2 - Combined NIC and FC cards l Shelf combinations are documented in the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. If DD Extended Retention is enabled on a DD860, the maximum usable storage capacity of an active tier is 144 TB. The retention tier can have a maximum usable capacity of 552 TB. The active and retention tiers have a total usable storage capacity of 696 TB. DD990
l 256 GB of RAM l 1 - NVRAM IO module (2 GB) l 4 - Quad-port SAS IO modules l 2 - 1 GbE ports on the motherboard l 0 to 4 - 1 GbE NIC IO cards for external connectivity l 0 to 3 - 10 GbE NIC cards for external connectivity l 0 to 3 - Dual-Port FC HBA cards for external connectivity l 0 to 3 - Combined NIC and FC cards, not to exceed three of any one specific IO module l Shelf combinations are documented in the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. If DD Extended Retention is enabled on a DD990, the maximum usable storage capacity of the active tier is 570 TB. The retention tier can have a maximum usable capacity of 1.4 PB. The active and retention tiers have a total usable storage capacity of 2 PB. DD4200
l 128 GB of RAM l 1 - NVRAM IO module (4 GB)
390 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
l 4 - Quad-port SAS IO modules l 1 - 1 GbE port on the motherboard l 0 to 6 - 1/10 GbE NIC cards for external connectivity l 0 to 6 - Dual-Port FC HBA cards for external connectivity l 0 to 6 - Combined NIC and FC cards, not to exceed four of any one specific IO module l Shelf combinations are documented in the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. If DD Extended Retention is enabled on a DD4200, the maximum usable storage capacity of the active tier is 192 TB. The retention tier can have a maximum usable capacity of 384 TB. The active and retention tiers have a total usable storage capacity of 576 TB. External connectivity is supported for DD Extended Retention configurations up to 16 shelves. DD4500 l 192 GB of RAM l 1 - NVRAM IO module (4 GB) l 4 - Quad-port SAS IO modules l 1 - 1 GbE port on the motherboard l 0 to 6 - 1/10 GbE NIC IO cards for external connectivity l 0 to 6 - Dual-Port FC HBA cards for external connectivity l 0 to 5 - Combined NIC and FC cards, not to exceed four of any one specific IO module l Shelf combinations are documented in the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. If DD Extended Retention is enabled on a DD4500, the maximum usable storage capacity of the active tier is 288 TB. The retention tier can have a maximum usable capacity of 864 TB. The active and retention tiers have a total usable storage capacity of 1.1 PB. External connectivity is supported for DD Extended Retention configurations up to 32 shelves. DD7200 l 128 GB of RAM for entry capacity; optional upgrade to 256 GB RAM for expanded capacity l 1 - NVRAM IO module (4 GB) l 4 - Quad-port SAS IO modules l 1 - 1 GbE port on the motherboard l 0 to 6 - 1/10 GbE NIC cards for external connectivity l 0 to 6 - Dual-Port FC HBA cards for external connectivity l 0 to 5 - Combined NIC and FC cards, not to exceed four of any one specific IO module l Shelf combinations are documented in the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. If DD Extended Retention is enabled on a DD7200, the maximum usable storage capacity of the active tier is 428 TB. The retention tier can have a maximum usable capacity of 1.2 PB. The active and retention tiers have a total usable storage capacity of 1.7 PB. External connectivity is supported for DD Extended Retention configurations up to 56 shelves. DD9500 l 256 GB of RAM for entry capacity; optional upgrade to 512 GB RAM for expanded capacity
Hardware supported for DD Extended Retention 391 DD Extended Retention
l 1 - NVRAM IO module (8 GB) l 4 - Quad-port SAS IO modules l 1 - Quad 1 GbE ports on the motherboard l 0 to 4 - 10 GbE NIC cards for external connectivity l 0 to 4 - Dual-Port 16 Gbe FC HBA cards for external connectivity l Shelf combinations are documented in the installation and setup guide for your DD system, and the expansion shelf hardware guides for your expansion shelves. If DD Extended Retention is enabled on a DD9500, the maximum usable storage capacity of the active tier is 864 TB. The retention tier can have a maximum usable capacity of 864 TB. The active and retention tiers have a total usable storage capacity of 1.7 PB. External connectivity is supported for DD Extended Retention configurations up to 56 shelves.
Licensing for DD Extended Retention DD Extended Retention is a licensed software option installed on a supported DD system. A separate shelf capacity license is needed for each storage shelf, for shelves installed in both the active tier and the retention tier. Shelf capacity licenses are specific to either an active or retention tier shelf. An Expanded-Storage license is required to expand the active tier storage capacity beyond the entry capacity, which varies by Data Domain model. You cannot use the additional storage without first applying the appropriate licenses.
Adding shelf capacity licenses for DD Extended Retention Every shelf in a DD Extended Retention-enabled DD system must have a separate license. Procedure
1. Select Administration > Licenses. 2. Click Add Licenses. 3. Enter one or more licenses, one per line, pressing the Enter key after each one. Click Add when you have finished. If there are any errors, a summary of the added licenses, and those not added because of the error, are listed. Select the erroneous License Key to fix it. Results The licenses for the DD system are displayed in two groups:
l Software option licenses, which are required for options such as DD Extended Retention and DD Boost. l Shelf Capacity Licenses, which display shelf capacity (in TiB), the shelf model (such as ES30), and the shelf’s storage tier (active or retention). To delete a license, select the license in the Licenses list, and click Delete Selected Licenses. If prompted to confirm, read the warning, and click OK to continue.
392 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
Configuring storage for DD Extended Retention Additional storage for DD Extended Retention requires the appropriate license or licenses and enough installed memory on the DD system to support it. Error messages display if more licenses or memory are needed. Procedure
1. Select Hardware > Storage tab. 2. In the Overview tab, select Configure Storage. 3. In the Configure Storage tab, select the storage to be added from the Available Storage list. 4. Select the appropriate Tier Configuration (or Active or Retention) from the menu. The active tier is analogous to a standard DD system and should be sized similarly. The maximum amount of storage that can be added to the active tier depends on the DD controller used. 5. Select the checkbox for the Shelf to be added. 6. Click the Add to Tier button. 7. Click OK to add the storage. 8. To remove an added shelf, select it in the Tier Configuration list, select Remove from Tier, and select OK.
Customer-provided infrastructure for DD Extended Retention Before enabling DD Extended Retention, your environment and setup must meet certain requirements.
l Specifications, site requirements, rack space, and interconnect cabling: See the EMC Data Domain Installation and Setup Guide for your DD system model. l Racking and cabling: It is recommended that you rack your system with future expansion in mind. All shelves are attached to a single DD system.
n See the EMC Data Domain Expansion Shelf Hardware Guide for your shelf model (ES20, ES30, or DS60).
Managing DD Extended Retention To set up and use DD Extended Retention on your DD system, you can use the DD System Manager and/or the DD CLI.
l The DD System Manager, formerly known as the Enterprise Manager, is a graphical user interface (GUI), which is described in this guide. l The archive commands, entered at the DD Command Line Interface (CLI) are described in the EMC Data Domain Operating System Command Reference Guide. The only command not available when you use the DD System Manager is archive report.
Configuring storage for DD Extended Retention 393 DD Extended Retention
Enabling DD systems for DD Extended Retention Before using a DD system for DD Extended Retention, you must have the correct license and the correct file system setup. Procedure 1. Ensure that the correct license is applied. Select Administration > Licenses, and check the Feature Licenses list for Extended Retention. 2. Select Data Management > File System > More Tasks > Enable DD Extended Retention. This option is available only if the file system has not already been configured for DD Extended Retention. Be aware that after DD Extended Retention has been enabled, it cannot be disabled without destroying the file system. a. If the file system is already enabled (as a non-DD Extended Retention system), you are prompted to disable it. Click Disable to do so. b. If prompted to confirm that you want to convert the file system for use by DD Extended Retention, click OK. After a file system is converted into a DD Extended Retention file system, the file system page is refreshed to include information about both tiers, and there is a new tab labeled Retention Units.
CLI Equivalent You can also verify that the Extended Retention license has been installed at the CLI: # license show ## License Key Feature ------1 AAAA-BBBB-CCCC-DDDD Replication 2 EEEE-FFFF-GGGG-HHHH VTL ------If the license is not present, each unit includes documentation – a quick install card – which shows the licenses that have been purchased. Enter the following command to populate the license key. # license add license-code Then, enable Extended Retention: # archive enable
Creating a two-tiered file system for DD Extended Retention DD Extended Retention has a two-tiered file system for the active and retention tiers. The DD system must have been enabled for DD Extended Retention before enabling this special file system. Procedure 1. Select Data Management > File System. 2. If a file system exists, destroy it. 3. Select More Tasks > Create file system. 4. Select a retention-capable file system and click Next. 5. Click Configure in the File System Create dialog box. Storage must be configured before the file system is created.
394 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
6. Use the Configure Storage dialog box to add and remove available storage from the Active and Retention Tiers, and click OK when you have finished. The storage in the active tier is used to create the active file system tier, and the storage in the retention tier is used to create a retention unit.
As of DD OS 5.5.1, only one retention unit per retention tier is allowed. However, systems set up prior to DD OS 5.5.1 may continue to have more than one retention unit, but you cannot add any more retention units to them.
7. Use the File System Create dialog box to: a. Select the size of the retention unit from the drop-down list. b. Select the Enable file system after creation option. c. Click Next. A Summary page shows the size of the active and retention tiers in the new file system.
8. Click Finish to create the file system. The progress of each creation step is shown, and a progress bar monitors overall status.
9. Click OK after the file system execution has completed. CLI Equivalent To add additional shelves, use this command once for each enclosure: # storage add tier archive enclosure 5 Create an archive unit, and add it to the file system. You are asked to specify the number of enclosures in the archive unit: # filesys archive unit add Verify that the archive unit is created and added to the file system: # filesys archive unit list all Check the file system, as seen by the system: # filesys show space
File system panel for DD Extended Retention After you have enabled a DD system for DD Extended Retention, the Data Management > File System panel will look slightly different (from a non-DD Extended Retention-enabled system). l State shows that the file system is either enabled or disabled. You can change the state by using the Disable/Enable button immediately to the right. l Clean Status shows the time the last cleaning operation finished, or the current cleaning status if the cleaning operation is currently running. If cleaning can be run, it shows a Start Cleaning button. When cleaning is running, the Start Cleaning button changes to a Stop Cleaning button. l Data Movement Status shows the time the last data movement finished. If data movement can be run, it shows a Start button. When data movement is running, the Start button changes to a Stop button.
File system panel for DD Extended Retention 395 DD Extended Retention
l Space Reclamation Status shows the amount of space reclaimed after deleting data in the retention tier. If space reclamation can be run, it shows a Start button. If it is already running, it shows Stop and Suspend buttons. If it was running previously and was suspended, it shows Stop and Resume buttons. There is also a More Information button that will display detailed information about starting and ending times, completion percentage, units reclaimed, space freed, etc. l Selecting More Tasks > Destroy lets you delete all data in the file system, including virtual tapes. This can be done only by a system administrator. l Selecting More Tasks > Fast Copy lets you clone files and MTrees of a source directory to a destination directory. Note that for DD Extended Retention-enabled systems, fast copy will not move data between the active and retention tiers. l Selecting More Tasks > Expand Capacity lets you expand the active or retention tier.
Expanding the active or retention tier When the file system is enabled, you can expand either the active or the retention tier. To expand the Active tier: Procedure 1. Select Data Management > File System > More Tasks > Expand Capacity. 2. In the Expand File System Capacity dialog, select Active Tier, then click Next. 3. Click Configure. 4. In the Configure Storage dialog, make sure that Active Tier is displayed as the Configure selection, and click OK. 5. After the configuration completes, you are returned to the Expand File System Capacity dialog. Select Finish to complete the active tier expansion.
To expand the retention tier: Procedure 1. Select Data Management > File System > More Tasks > Expand Capacity. 2. In the Expand File System Capacity dialog, select Retention Tier, then select Next. 3. If a retention unit is available, you will see the Select Retention Unit dialog. Select the retention unit you want to expand and then Next. If a retention unit is not available, you will see the Create Retention Unit dialog, and you must create a retention unit before proceeding.
To ensure optimal performance of a DD system with DD Extended Retention enabled, you should always expand the retention tier in at least two-shelf increments. You should also not wait until the retention unit is nearly full before expanding it.
4. Select the size to expand the retention unit, then click Configure. 5. After configuration completes, you are returned to the Expand File System Capacity dialog. Click Finish to complete the retention tier expansion.
396 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
Reclaiming space in the retention tier You can reclaim space from deleted data in the retention tier by running space reclamation (introduced in DD OS 5.3). Space reclamation also occurs during file system cleaning. Procedure
1. Select Data Management > File System. Just above the tabs, Space Reclamation Status shows the amount of space that is reclaimed after deleting data in the retention tier. 2. If space reclamation can be run, it shows a Start button. If it is already running, it shows Stop and Suspend buttons. If it was running previously and was suspended, it shows Stop and Resume buttons. 3. Click More Information for details about the cycle name, start and end times, effective run time, percent completed (if in progress), units reclaimed, space freed on target unit, and total space freed.
When you use the archive space-reclamation command, the system runs space-reclamation in the background until it is manually stopped unless you use the one-cycle option. You can also use the archive space-reclamation schedule set command to set the starting time for space-reclamation.
CLI Equivalent To enable space reclamation: # archive space-reclamation start To disable space reclamation: # archive space-reclamation stop To show the status of space reclamation: # archive space-reclamation status-detailed Space-reclamation will start when 'archive data-movement' completes.
Previous Cycle: ------Start time : Feb 21 2014 14:17 End time : Feb 21 2014 14:49 Effective run time : 0 days, 00:32. Percent completed : 00 % (was stopped by user) Units reclaimed : None Space freed on target unit : None Total space freed : None
File system tabs for DD Extended Retention After you have enabled a DD system for DD Extended Retention, the Data Management > File System tabs will also look slightly different (from a non-DD Extended Retention- enabled system), and there will be one additional tab: Retention Units Summary Tab The Summary tab displays information about disk space usage and compression for both the active and retention tiers.
File system tabs for DD Extended Retention 397 DD Extended Retention
Space Usage: Shows the total size, amount of space used, and amount of available space and combined totals for active and retention tiers. The amount of cleanable space is shown for the active tier. Active Tier and Retention Tier: Shows the pre-compression and post-compression values currently used and those written in the last 24 hours. Also shows the global, local, and total compression (reduction percentage) factors. Retention Units Tab The Retention Units tab displays the retention unit(s). As of DD OS 5.5.1.4, only one retention unit per retention tier is allowed. However, systems set up prior to DD OS 5.5.1.4 may continue to have more than one retention unit, but you will not be allowed to add any more retention units to them. The following information is displayed: the unit’s State (New, Empty, Sealed, Target, or Cleaning), its Status (Disabled, Ready, or Stand-by), its Start Date (when it was moved to the retention tier), and the Unit Size. The unit will be in the cleaning state if space reclamation is running. If the unit has been sealed, meaning no more data can be added, the Sealed Date is provided. Selecting the retention unit's checkbox displays additional information (Size, Used, Available, and Cleanable) in the Detailed Information panel. There are two buttons: Delete (for deleting the unit) and Expand (for adding storage to a unit). The unit must be in a new or target state to be expanded. Configuration Tab The Configuration Tab lets you configure your system. Selecting the Options Edit button displays the Modify Settings dialog, where you can change Local Compression Type [options are none, lz (the default), gz, and gzfast] and Retention Tier Local Comp(ression) [options are none, lz, gz (the default), and gzfast], as well as enable Report Replica Writable. Selecting the Clean Schedule Edit button displays the Modify Schedule dialog, where you can change the cleaning schedule, as well as the throttle percentage. Selecting the Data Movement Policy Edit button displays the Data Movement Policy dialog, where you can set several parameters. File Age Threshold is a system-wide default that applies to all MTrees for which you have not set a custom default. The minimum value is 14 days. Data Movement Schedule lets you establish how often data movement will be done; the recommended schedule is every two weeks. File System Cleaning lets you elect not to have a system cleaning after data movement; however, it is strongly recommended that you leave this option selected. File Age Threshold per MTree Link Selecting the File Age Threshold per MTree link will take you from the File System to the MTree area (also accessible by selecting Data Management > MTree), where you can set a customized File Age Threshold for each of your MTrees. Select the MTree, and then select Edit next to Data Movement Policy. In the Modify Age Threshold dialog, enter a new value for File Age Threshold, and select OK. As of DD OS 5.5.1, the minimum value is 14 days. Encryption Tab The Encryption tab lets you enable or disable Encryption of Data at Rest, which is supported only for systems with a single retention unit. As of 5.5.1, DD Extended Retention supports only a single retention unit, so systems set up during, or after, 5.5.1 will have no problem complying with this restriction. However, systems set up prior to 5.5.1 may have more than one retention unit, but they will not work with Encryption of Data at Rest until all but one retention unit has been removed, or data has been moved or migrated to one retention unit.
398 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
Space Usage Tab The Space Usage Tab lets you select one of three chart types [(entire) File System; Active (tier); Archive (tier)] to view space usage over time in MiB. You can also select a duration value (7, 30, 60, or 120 days) at the upper right. The data is presented (color-coded) as pre-compression written (blue), post-compression used (red), and the compression factor (black). Consumption Tab The Consumption Tab lets you select one of three chart types [(entire) File System; Active (tier); Archive (tier)] to view the amount of post-compression storage used and the compression ratio over time, which enables you to view consumption trends. You can also select a duration value (7, 30, 60, or 120 days) at the upper right. The Capacity checkbox lets you choose whether to display the post-compression storage against total system capacity. Daily Written Tab The Daily Written Tab lets you select a duration (7, 30, 60, or 120 days) to see the amount of data written per day. The data is presented (color-coded) in both graph and table format as pre-compression written (blue), post-compression used (red), and the compression factor (black).
Expanding a retention unit To ensure optimal performance, do not wait until a retention unit is nearly full before expanding it, and do not expand it in 1-shelf increments. Storage cannot be moved from the active tier to the retention tier after the file system has been created. Only unused enclosures can be added to the retention tier. Procedure
1. Select Data Management > File System > Retention Units. 2. Select the retention unit. Note that if cleaning is running, a retention unit cannot be expanded. 3. Click Expand. The system displays the current retention tier size, an estimated expansion size, and a total expanded capacity. If additional storage is available you can click the Configure link. 4. Click Next. The system displays a warning telling you that you cannot revert the file system to its original size after this operation. 5. Click Expand to expand the file system.
Deleting a retention unit If all of the files on a retention unit are no longer needed, deleting them makes the unit available for reuse. You can generate a file location report to make sure that the retention unit is indeed empty, delete the retention unit, and then add it as a new retention unit. Procedure
1. Select Data Management > File System and click Disable to disable the file system if it is running. 2. Select Data Management > File System > Retention Units. 3. Select the retention unit.
File system tabs for DD Extended Retention 399 DD Extended Retention
Modifying retention tier local compression You can modify the local compression algorithm for subsequent data movement to the retention tier. Procedure 1. Select Data Management > File System > Configuration. 2. Click Edit to the right of Options. 3. Select one of the compression options from the Retention Tier Local Comp menu, and click OK. The default is gz, which is a zip-style compression that uses the least amount of space for data storage (10% to 20% less than lz on average; however, some data sets achieve much higher compression).
Understanding the Data Movement Policy A file is moved from the active to the retention tier based on the date it was last modified. For data integrity, the entire file is moved at this time. The Data Movement Policy establishes two things: a File Age Threshold and a Data Movement Schedule. If data has not changed during the period of days set by the File Age Threshold, it is moved from the active to the retention tier on the date established by the Data Movement Schedule.
As of DD OS 5.5.1, the File Age Threshold must be a minimum of 14 days.
You can specify different File Age Thresholds for each defined MTree. An MTree is a subtree within the namespace that is a logical set of data for management purposes. For example, you might place financial data, emails, and engineering data in separate MTrees. To take advantage of the space reclamation feature, introduced in DD OS 5.3, it is recommended that you schedule data movement and file system cleaning on a bi-weekly (every 14 days) basis. By default, cleaning is always run after data movement completes. It is highly recommended that you do not change this default. Avoid these common sizing errors: l Setting a Data Movement Policy that is overly aggressive; data will be moved too soon. l Setting a Data Movement Policy that is too conservative: after the active tier fills up, you will not be able to write data to the system. l Having an undersized active tier and then setting an overly aggressive Data Movement Policy to compensate. Be aware of the following caveats related to snapshots and file system cleaning: l Files in snapshots are not cleaned, even after they have been moved to the retention tier. Space cannot be reclaimed until the snapshots have been deleted. l It is recommended that you set the File Age Threshold for snapshots to the minimum of 14 days. Here are two examples of how to set up a Data Movement Policy. l You could segregate data with different degrees of change into two different MTrees and set the File Age Threshold to move data soon after the data stabilizes. Create MTree A for daily incremental backups and MTree B for weekly fulls. Set the File Age
400 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
Threshold for MTree A so that its data is never moved, but set the File Age Threshold for MTree B to 14 days (the minimum threshold). l For data that cannot be separated into different MTrees, you could do the following. Suppose the retention period of daily incremental backups is eight weeks, and the retention period of weekly fulls is three years. In this case, it would be best to set the File Age Threshold to nine weeks. If it were set lower, you would be moving daily incremental data that was actually soon to be deleted. Modifying the Data Movement Policy You can set different Data Movement Policies for each MTree. Procedure
1. Select Data Management > File System > Configuration. 2. Click Edit to the right of Data Movement Policy. 3. In the Data Movement Policy dialog, specify the system-wide default File Age Threshold value in number of days. As of DD OS 5.5.1, this value must be greater than or equal to 14 days. This value applies to newly created MTrees and MTrees that have not been assigned a per-MTree age threshold value using the File Age Threshold per MTree link (see step 7). When data movement starts, all files that have not been modified for the specified threshold number of days will be moved from the active to the retention tier. 4. Specify a Data Movement Schedule, that is, when data movement should take place; for example, daily, weekly, bi-weekly (every 14 days), monthly, or on the last day of the month. You can also pick a specific day or days, and a time in hours and minutes. It is highly recommended that you schedule data movement and file system cleaning on a bi-weekly (every 14 days) basis, to take advantage of the space reclamation feature (introduced in DD OS 5.3). 5. Specify a Data Movement Throttle, that is, the percentage of available resources the system uses for data movement. A value of 100% indicates that data movement will not be throttled. 6. By default, file system cleaning is always run after data movement completes. It is highly recommended that you leave Start file system clean after Data Movement selected. 7. Select OK. 8. Back in the Configuration tab, you can specify age threshold values for individual MTrees by using the File Age Threshold per MTree link at the lower right corner.
CLI Equivalent To set the age threshold: # archive data-movement policy set age-threshold
File system tabs for DD Extended Retention 401 DD Extended Retention
l days sun time 00:00 l days mon,tue time 00:00 l days 2 time 10:00 l days 2,15 time 10:00 l days last time 10:00 - last day of the month To verify the migration schedule: # archive data-movement schedule show To disable the file cleaning schedule:
The reason for disabling the cleaning schedule is to eliminate a scheduling conflict between cleaning and data movement. At the conclusion of data movement, cleaning will automatically start. If you disable data movement, you should re-enable file system cleaning.
# filesys clean set schedule never
Starting or stopping data movement on demand Even when you have a regular Data Movement Policy, you can also start or stop data movement on demand. Procedure
1. Select Data Management > File System. 2. Click Start to the right of Data Movement Status. 3. The Start Data Movement dialog warns that data is to be moved from the active to the retention tier, as defined by your Data Movement Policy, followed by a file system cleaning. Select Start to start the data movement. If a file system cleaning happens to already be in progress, data movement will occur after that cleaning completes. However, another cleaning will be automatically started after this on-demand data movement completes, as well. 4. The Start button will be replaced by a Stop button. 5. At any time, if you want to stop data movement, click Stop and click OK in the Stop Data Movement dialog to confirm. Using data movement packing Data is compacted in the target partition after every file migration (as of DD OS 5.2). By default, this feature, which is called data movement packing, is enabled. When this feature is enabled, the overall compression of the retention tier improves, but there is a slight increase in migration time. To determine if this feature is enabled, select Data Management > File System > Configuration. The current value for Packing data during Retention Tier data movement can be either Enabled or Disabled. Consult with a system engineer to change this setting.
402 EMC Data Domain Operating System 5.7 Administration Guide DD Extended Retention
Upgrades and recovery with DD Extended Retention The following sections describe how to perform software and hardware upgrades, and how to recover data, for DD Extended Retention-enabled DD systems.
Upgrading to DD OS 5.7 with DD Extended Retention The upgrade policy for a DD Extended Retention-enabled DD system is the same as for a standard DD system. Upgrading from up to two major prior releases is supported. For instructions on how to upgrade the DD OS, refer to the upgrade instructions section of the Release Notes for the target DD OS version. When upgrading a DD Extended Retention-enabled DD system to DD OS 5.7, be sure to update existing data movement schedules to bi-weekly (14 days) to take advantage of the space reclamation feature. DD Extended Retention-enabled DD systems automatically run cleaning after data movement completes; therefore, do not schedule cleaning separately using the DD System Manager or CLI (command line interface). If the active tier is available, the process upgrades the active tier and the retention unit, and puts the system into a state that the previous upgrade has not been verified to be complete. This state is cleared by the file system after the file system is enabled and has verified that the retention tier has been upgraded. A subsequent upgrade is not permitted until this state is cleared. If the active tier is not available, the upgrade process upgrades the system chassis and places it into a state where it is ready to create or accept a file system. If the retention unit becomes available after the upgrade process has finished, the unit is automatically upgraded when it is plugged into the system, or at the next system start.
Upgrading hardware with DD Extended Retention You can upgrade a DD Extended Retention-enabled DD system to a later or higher performance DD Extended Retention-enabled DD system. For example, you could replace a DD Extended Retention-enabled DD860 with a DD Extended Retention-enabled DD990
Consult your contracted service provider, and refer to the instructions in the appropriate System Controller Upgrade Guide.
This type of upgrade affects DD Extended Retention as follows:
l If the new system has a more recent version of DD OS than the active and retention tiers, the active and retention tiers are upgraded to the new system's version. Otherwise, the new system is upgraded to the version of the active and retention tiers. l The active and retention tiers that are connected to the new system become owned by the new system. l If there is an active tier, the registry in the active tier is installed in the new system. Otherwise, the registry in the retention tier with the most recently updated registry is installed in the new system.
Upgrades and recovery with DD Extended Retention 403 DD Extended Retention
Recovering a DD Extended Retention-enabled system If the active tier, and a subset of the retention units are lost, on a DD Extended Retention- enabled DD system, and there is no replica available, EMC Support may be able to reconstitute any remaining sealed retention units into a new DD system. A DD Extended Retention-enabled DD system is designed to remain available to service read and write requests when one or more retention units are lost. The file system may not detect that a retention unit is lost until the file system restarts or tries to access data stored in the retention unit. The latter circumstance may trigger a file system restart. After the file system has detected that a retention unit is lost, it returns an error in response to requests for data stored in that unit. If the lost data cannot be recovered from a replica, EMC Support might be able to clean up the system by deleting the lost retention unit and any files that reside fully or partially in it.
Using replication recovery The replication recovery procedure for a DD Extended Retention-enabled DD system depends on the replication type
l Collection replication – The new source must be configured as a DD Extended Retention-enabled DD system with the same number (or more) of retention units as the destination. The file system must not be enabled on the new source until the retention units have been added, and replication recovery has been initiated.
If you need to recover only a portion of a system, such as one retention unit, from a collection replica, contact EMC Support.
l MTree replication – See the MTree Replication section in the Working with DD Replicator chapter. l DD Boost managed file replication – See the EMC Data Domain Boost for OpenStorage Administration Guide.
Recovering from system failures A DD Extended Retention-enabled DD system is equipped with tools to address failures in different parts of the system. Procedure 1. Restore the connection between the system controller and the storage. If the system controller is lost, replace it with a new system controller. 2. If there is loss of data and a replica is available, try to recover the data from the replica. If a replica is not available, limit any loss of data by leveraging the fault isolation features of DD Extended Retention through EMC Support.
404 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 17 DD Retention Lock
This chapter includes:
l DD Retention Lock overview. 406 l Supported data access protocols. 408 l Enabling DD Retention Lock on an MTree. 409 l Client-Side Retention Lock file control. 412 l System behavior with DD Retention Lock. 417
DD Retention Lock 405 DD Retention Lock
DD Retention Lock overview When data is locked on an MTree that is enabled with the DD Retention Lock software option, DD Retention Lock helps ensure that data integrity is maintained. Any data that is locked cannot be overwritten, modified, or deleted for a user-defined retention period of up to 70 years. There are two DD Retention Lock editions:
l EMC Data Domain Retention Lock Governance Edition retains the functionality of Data Domain Retention Lock software prior to DD OS 5.2. You can use Data Domain Retention Lock Governance to define retention policies on data that is to be retained for a specific period of time to meet internal IT governance policies implemented by the system administrator. l EMC Data Domain Retention Lock Compliance Edition enables you to meet the strictest data permanence requirements of regulatory standards, such as those of SEC 17a-4(f). The full list of regulatory standards includes:
n CFTC Rule 1.31b n FDA 21 CFR Part 11 n Sarbanes-Oxley Act n IRS 98025 and 97-22 n ISO Standard 15489-1 n MoREQ2010 For certification information, see Compliance Assessments - Summary and Conclusions – EMC Data Domain Retention Lock Compliance Edition at: https://www.emc.com/collateral/analyst-reports/cohasset-dd-retention-lock-assoc- comp-assess-summ-ar.pdf (Login is required.) Compliance with these standards ensures that files locked on a Data Domain system using Data Domain Retention Lock Compliance Edition software cannot be altered or destroyed before the retention period expires. Data Domain Retention Lock Compliance Edition requires a security officer for implementation of policies. An audit log file is accessible by the administrator or security officer. Each edition requires a separate, add-on license, and either or both can be used on a single Data Domain system. The retention-locking protocol is the same for both the DD Retention Lock Governance and Compliance Editions. The differences in use stem from the system behavior for the DD Retention Lock Compliance Edition, since it places strict restrictions to meet compliance requirements. For an overview, see the EMC Data Domain Retention Lock Software – A Detailed Review (a white paper) available at: https://www.emc.com/collateral/hardware/white-papers/h10666-data-domain- retention-lock-wp.pdf (Login is required.) The DD Retention Lock Governance Edition does not require a security officer and provides a higher degree of flexibility for archive data retention on Data Domain systems. For archive compliance storage requirements, SEC rules require that a separate copy of retention-locked data must be stored with the same retention requirements as the original. Retention-locked files can be replicated using DD Replicator software to another
406 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
Data Domain system. If a retention-locked file is replicated, it remains retention locked on the destination system, with the same level of protection as the source file. The topics that follow provide additional information on the DD Retention Lock software option.
DD Retention Lock protocol Only files that are explicitly committed to be retention-locked files are retention locked on the Data Domain system. Files are committed to be retention-locked files through client- side file commands issued while DD Retention Lock Governance or Compliance is enabled on the MTree containing the files.
Linux, Unix, and Windows client environments are supported.
Files that are written to shares or exports that are not committed to be retained (even if DD Retention Lock Governance or Compliance is enabled on the MTree containing the files) can be modified or deleted at any time. Retention locking prevents any modification or deletion of files under retention from occurring directly from CIFS shares or NFS exports during the retention period specified by a client-side atime update command. Some archive applications and backup applications can issue this command when appropriately configured. Applications or utilities that do not issue this command cannot lock files using DD Retention Lock. Retention-locked files are always protected from modification and premature deletion, even if retention locking is subsequently disabled or if the retention-lock license is no longer valid. You cannot rename or delete non-empty folders or directories within an MTree that is retention-lock enabled. However, you can rename or delete empty folders or directories and create new ones. The retention period of a retention-locked file can be extended (but not reduced) by updating the file’s atime. For both DD Retention Lock Governance and Compliance, once the retention period for a file expires, the file can be deleted using a client-side command, script, or application. However, the file cannot be modified even after the retention period for the file expires. The Data Domain system never automatically deletes a file when its retention period expires.
DD Retention Lock flow The general flow of activities with DD Retention Lock. 1. Enable MTrees for DD Retention Lock Governance or Compliance retention locking using the DD System Manager or DD OS commands issued from the system console. 2. Commit files to be retention locked on the Data Domain system using client-side commands issued by an appropriately configured archiving or backup application, manually, or via scripts.
Windows clients may need to download utility programs for DD OS compatibility.
3. Optionally, extend file retention times using client-side commands. 4. Optionally, delete files with expired retention periods using client-side commands.
DD Retention Lock protocol 407 DD Retention Lock
Supported data access protocols DD Retention Lock is compatible with industry-standard, NAS-based Write-Once-Read- Many (WORM) protocols, and integration is qualified with archive applications such as Symantec Enterprise Vault, EMC SourceOne, EMC Cloud Tiering Appliance, EMC DiskXtender, and so on. Customers using backup applications such as CommVault can also develop custom scripts to use the EMC Data Domain Retention Lock software option. To check whether an archive application is tested and certified for DD Retention Lock, refer to the EMC Data Domain Archive Application Compatibility Guide. The protocol support of DD Retention Lock software is as follows:
l NFS is supported with both DD Retention Lock Governance and Compliance. l CIFS is supported with both DD Retention Lock Governance and Compliance. l VTL is supported with DD Retention Lock Governance, but not with DD Retention Lock Compliance. Virtual tapes, here referred to as tapes, are represented as files on the Data Domain file system.
n When you create a storage pool, a collection of tapes that map to a directory on the file system, you are creating an MTree, unless you specifically select to create the older style directory pool (for backward compatibility). You can also convert storage pools created prior to DD OS 5.3 to MTrees. These MTrees can be retention locked and replicated. n You can retention-lock one or more tapes using the vtl tape modify command, described in the EMC Data Domain Operating System Command Reference Guide. The mtree retention-lock revert path command can be used to revert the retention-locked state of tapes locked with the vtl tape modify command. After the tape is unlocked, updates can be made to it. The unlocked state won't be visible via the DD System Manager or CLI until the VTL service is disabled then enabled; however, updates will be applied to the unlocked tape. This capability is only for the DD Retention Lock Governance Edition. n The retention time for tapes can be displayed using the vtl tape show command with the time-display retention argument.
n You can retention-lock an individual tape using the DD System Manager.
408 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
l DD Boost is supported with both DD Retention Lock Governance and Compliance. If client-side scripts are used to retention-lock backup files or backup images, and if a backup application (Symantec NetBackup, for example) is also used on the system via DD Boost, be aware that the backup application may not share the context of the client-side scripts. Thus, when a backup application attempts to expire or delete files that were retention locked via the client-side scripts, space is not released on the EMC Data Domain system. Data Domain recommends that administrators change their retention period policy to align with the retention lock time. This applies to all the backup applications that are integrated with DD Boost: Avamar, Symantec NetBackup, Symantec Backup Exec, EMC NetWorker, and so on.
Enabling DD Retention Lock on an MTree Only files within DD Retention Lock Governance or Compliance enabled MTrees can be retention-locked. MTrees enabled for DD Retention Lock Compliance cannot be converted to DD Retention Lock Governance MTrees and vice versa. The procedures that follow show how to enable MTrees for either DD Retention Lock Governance or DD Retention Lock Compliance.
Enabling DD Retention Lock Governance on an MTree Add a DD Retention Lock Governance license to a system, and then enable DD Retention Lock Governance on one or more MTrees. Procedure 1. Add the Retention Lock Governance license, if it is not listed under Feature Licenses.
a. Select Administration > Licenses b. In the Licenses area click Add Licenses. c. In the License Key text box, type the license key.
License keys are case-insensitive. Include the hyphens when typing keys.
d. Click Add. 2. Select an MTree for retention locking.
a. Select Data Management > MTree. b. Select the MTree you want to use for retention locking. You can also create an empty MTree and add files to it later. 3. Click the MTree Summary tab to display information for the selected MTree. 4. Scroll down to Retention Lock area and click Edit to the right of Retention Lock. 5. Enable DD Retention Lock Governance on the MTree and change the default minimum and maximum retention lock periods for the MTree, if required. Perform the following actions in the Modify Retention Lock dialog box:
a. Select Enable to enable DD Retention Lock Governance on the MTree.
Enabling DD Retention Lock on an MTree 409 DD Retention Lock
b. To change the minimum or maximum retention period for the MTree, modify the minimum or maximum period:
Type a number for the interval in the text box (for example, 5 or 14). From the drop-down list, select an interval (minutes, hours, days, years).
Specifying a minimum retention period of less than 12 hours, or a maximum retention period longer than 70 years, results in an error.
c. Click OK to save the settings. After you close the Modify Retention Lock dialog box, which is updated MTree information appears in the Retention Lock area. 6. Check retention lock information for the MTree. Note the following retention lock fields:
n The Status field indicates the read/write access for the MTree, the type of retention locking on the MTree, and whether retention locking is enabled or disabled. l Bottom:
n The Status field indicates whether retention locking is enabled for the MTree. n The Retention Period field indicates minimum and maximum retention periods for the MTree. The retention period that is specified for a file in the MTree must be equal to or greater than the minimum retention period and equal to or less than the maximum retention period. n The UUID field is a unique identification number that is generated for the MTree.
To check retention lock configuration settings for any MTree, select the MTree in the Navigation Panel, then click the Summary tab.
After you finish Retention-lock files in a retention-lock-enabled MTree.
Enabling DD Retention Lock Compliance on an MTree Add a DD Retention Lock Compliance license to a system, set up a system administrator and one or more security officers, configure and enable the system to use DD Retention Lock Compliance software, and then enable DD Retention Lock Compliance on one or more MTrees. Procedure 1. Add the DD Retention Lock Compliance license on the system, if it is not present. a. First, check whether the license is already installed. license show b. If the RETENTION-LOCK-COMPLIANCE feature is not displayed, install the license.
410 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
license add license-key
License keys are case-insensitive. Include the hyphens when typing keys.
2. Set up one or more security officer users accounts according to Role-Base Access Control (RBAC) rules. a. In the system administrator role, add a security officer account. user add user role security b. Enable the security officer authorization. authorization policy set security-officer enabled 3. Configure and enable the system to use DD Retention Lock Compliance.
Enabling DD Retention Lock Compliance enforces many restrictions on low-level access to system functions used during troubleshooting. Once enabled, the only way to disable DD Retention Lock Compliance is to initialize and reload the system, which results in destroying all data on the system.
a. Configure the system to use DD Retention Lock Compliance. system retention-lock compliance configure The system automatically reboots.
b. After the restart process is complete, enable DD Retention Lock Compliance on the system. system retention-lock compliance enable 4. Enable compliance on an MTree that will contain retention-locked files. mtree retention-lock enable mode compliance mtree mtree-path
Compliance cannot be enabled on /backup or pool MTrees.
5. To change the default minimum and maximum retention lock periods for a compliance-enabled MTree, type the following commands with security officer authorization.
mtree retention-lock set min-retention-period period mtree mtree-path mtree retention-lock set max-retention-period period mtree mtree-path
The retention period is specified in the format [number] [unit]. For example: 1 min, 1 hr, 1 day, 1 mo, or 1 year. Specifying a minimum retention period of less than 12 hours, or a maximum retention period longer than 70 years, results in an error.
Repeat steps 4 and 5 to enable additional MTrees.
Enabling DD Retention Lock Compliance on an MTree 411 DD Retention Lock
After you finish Retention lock files reside in a retention-lock-enabled MTree.
Client-Side Retention Lock file control This section describes the DD Retention Lock client command interface for locking files stored on EMC Data Domain systems. Client commands are the same for DD Retention Lock Governance and Compliance. Linux, Unix, and Windows client environments are supported; however, Windows clients may need to download utility programs with commands to lock files.
If your application already supports industry-standard WORM, writing a WORM file to a DD Retention Lock Governance or Compliance enabled MTree will lock the file on the Data Domain system. The retention time in the application should agree with the DD Retention Lock settings. You do not need to use the commands described in this section. To check whether an application is tested and certified for the DD Retention Lock, refer to the EMC Data Domain Archive Application Compatibility Guide.
Some client machines using NFS, but running a legacy OS, cannot set retention time later than 2038. The NFS protocol doesn’t impose the 2038 limit and allows to specifying times until 2106. Further, DD OS doesn’t impose the 2038 limit.
Client-side commands are used to manage the retention locking of individual files. These commands apply to all retention-lock-capable Data Domain systems and must be issued in addition to the setup and configuration of the DD Retention Lock software on the Data Domain system. Required Tools for Windows Clients You need the touch.exe command to perform retention-locking from a Windows-based client. To obtain this command, download and install utilities for Linux/Unix-based applications according to your Windows version. These utilities are best recommendations from EMC and should be used per customer environment.
l For Windows 8, Windows 7, Windows Server 2008, Windows Vista, Windows Server 2003, and Windows XP: http://sourceforge.net/projects/unxutils/files/latest l For Windows Server 2008, Windows Vista Enterprise, Windows Vista Enterprise 64-bit edition, Windows Vista SP1, Windows Vista Ultimate, and Windows Vista Ultimate 64- bit edition: http://www.microsoft.com/download/en/details.aspx?displaylang=en&id=23754 l For Windows Server 2003 SP1 and Windows Server 2003 R2: http://www.microsoft.com/download/en/details.aspx?displaylang=en&id=20983
The touch command for Windows may have a different format than the Linux examples in this chapter.
Follow the installation instructions provided and set the search path as needed on the client machine.
412 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
Client Access to Data Domain System Files After an MTree is enabled for DD Retention Lock Governance or Compliance, you can:
l Create a CIFS share based on the MTree. This CIFS share can be used on a client machine. l Create an NFS mount for the MTree and access its files from the NFS mount point on a client machine.
The commands listed in this section are to be used only on the client. They cannot be issued through the DD System Manager or CLI. Command syntax may vary slightly, depending on the utility you are using.
The topics that follow describe how to manage client-side retention lock file control.
Setting Retention Locking on a file To perform retention locking on a file, change the last access time (atime) of the file to the desired retention time of the file, that is, the time when the file can be deleted. This action is usually performed using the archive application, and all the archive applications that are qualified on Data Domain systems today (per the EMC Data Domain Archive Application Compatibility Guide) follow the basic locking protocol outlined here. The future atime you specify must respect the minimum and maximum retention periods of the file’s MTree (as offsets from the current time), as shown in the next figure.
Setting Retention Locking on a file 413 DD Retention Lock
Figure 26 Valid and invalid atimes for retention locking files
Some client machines using NFS, but running a legacy OS, cannot set retention time later than 2038. The NFS protocol doesn’t impose the 2038 limit and allows to specifying times until 2106. Further, DD OS doesn’t impose the 2038 limit.
Errors are permission-denied errors (referred to as EACCESS, a standard POSIX error). These are returned to the script or archive application setting the atime.
A file must be completely written to the Data Domain system before it is committed to be a retention-locked file.
The following command can be used on clients to set the atime: touch -a -t [atime] [filename] The format of atime is: [[YY]YY] MMDDhhmm[.ss] For example, suppose the current date and time is 1 p.m. on January 18, 2012 (that is, 201201181300), and the minimum retention period is 12 hours. Adding the minimum retention period of 12 hours to that date and time results in a value of 201201190100. Therefore, if the atime for a file is set to a value greater than 201201190100, that file becomes retention locked.
414 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
The following command: ClientOS# touch -a -t 201412312230 SavedData.dat will lock file SavedData.dat until 10:30 p.m. December 31, 2014.
Extending Retention Locking on a file To extend the retention time of a retention-locked file, set the file’s atime to a value greater than the file’s current atime but less than the maximum retention period of the file’s MTree (as an offset from the current time), as shown in the next figure. Figure 27 Valid and invalid atimes for extending retention locking on files
For example, changing the atime from 201412312230 to 202012121230 using the following command: ClientOS# touch -a -t 202012121230 SavedData.dat will cause the file to be locked until 12:30 p.m. December 12, 2020.
Extending Retention Locking on a file 415 DD Retention Lock
Some client machines using NFS, but running a very old OS, cannot set retention time later than 2038. The NFS protocol doesn’t impose the 2038 limit and allows to specifying times until 2106. Further, DD OS doesn’t impose the 2038 limit.
Errors are permission-denied errors (referred to as EACCESS, a standard POSIX error). These are returned to the script or archive application setting the atime.
Identifying a Retention-Locked file The atime value for a retention-locked file is its retention time. To determine whether a file is retention locked, try to set the atime of the file to a value earlier than its current atime. This action will fail with a permission-denied error if and only if the file is a retention- locked file. First, list the current atime value, and then execute the touch command with an earlier atime using these commands:
ls -l --time=atime [filename] touch -a -t [atime] [filename]
The following example shows the command sequence: ClientOS# ls -l --time=atime SavedData.dat 202012121230 ClientOS# touch -a -t 202012111230 SavedData.dat If the atime of SavedData.dat is 202012121230 (12:30 p.m. December 12, 2020) and the touch command specifies an earlier atime, 202012111230 (12:30 p.m. December 11, 2020), the touch command fails, indicating that SavedData.dat is retention- locked.
The --time=atime option is not supported in all versions of Unix.
Specifying a directory and touching only those files Use the command line to create a root directory containing the files for which access times will change. In this routine, root directory to start from contains the files on which you want to change access times using this client system command: find [root directory to start from] -exec touch -a -t [expiration time] <> \; For example: ClientOS# find [/backup/data1/] -exec touch -a -t 202012121230 <> \;
Reading a list of files and touching only those files In this routine, name of file list is the name of a text file that contains the names of the files on which you want to change access times. Each line contains the name of one file. Here is the client system command syntax: touch -a -t [expiration time] 'cat [name of file list]' For example:
416 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
ClientOS# touch -a -t 202012121230 ‘cat /backup/data1/filelist.txt‘
Deleting or expiring a file Delete or expire a file with an expired retention lock using a client application, or delete a file using a standard file-delete command. Expiring a file using an application makes the file inaccessible to the application. The file may or may not actually be removed from the Data Domain system by the expiration operation. If it is not removed, the application often provides a separate delete operation. You must have the appropriate access rights to delete the file, independent of the DD Retention Lock software.
If the retention period of the retention-locked file has not expired, the delete operation results in a permission-denied error.
Privileged delete For Retention Lock Governance (only), you can delete retention locked files using this two step process. Procedure 1. Use the mtree retention-lock revert path command to revert the retention locked file. 2. Delete the file on the client system using the rm filename command.
Using ctime or mtime on Retention-Locked files ctime is the last-metadata-change time of a file. ctime ctimegets set to the current time when any of the follow events occur:
l A non-retention-locked file is retention locked. l The retention time of a retention-locked file is extended. l A retention-locked file is reverted.
User access permissions for a retention-locked file are updated using the Linux command line tool chmod.
mtime mtime is the last-modified time of a file. It changes only when the contents of the file change. So, the mtime of a retention-locked file cannot change.
System behavior with DD Retention Lock System behavior topics are discussed separately for DD Retention Lock Governance and DD Retention Lock Compliance in the sections that follow.
Deleting or expiring a file 417 DD Retention Lock
DD Retention Lock governance Certain DD OS commands behave differently when using DD Retention Lock Governance. The following sections describe the differences for each.
Replication Collection replication, MTree replication, and directory replication replicate the locked or unlocked state of files. Files that are governance retention locked on the source are governance retention locked on the destination and have the same level of protection. For replication, the source system must have a DD Retention Lock Governance license installed—a license is not required on the destination system. Replication is supported between systems that are:
l Running the same major DD OS version (for example, both systems are running DD OS 5.5.x.x). l Running DD OS versions within the next two consecutive higher or lower major releases (for example, 5.3.x.x to 5.5.x.x or 5.5.x.x to 5.3.x.x). Cross-release replication is supported only for directory and MTree replication.
MTree replication is not supported for DD OS 5.0 and earlier.
l Collection replication and MTree replication replicate the minimum and maximum retention periods configured on MTrees to the destination system. l Directory replication does not replicate the minimum and maximum retention periods to the destination system. The procedure for configuring and using collection, MTree, and directory replication is the same as for Data Domain systems that do not have a DD Retention Lock Governance license. Replication Resync The replication resync destination command tries to bring the destination into sync with the source when the MTree or directory replication context is broken between destination and source systems. This command cannot be used with collection replication. Note that:
l If the destination directory has retention lock enabled, but the source directory does not have retention lock enabled, then a resync of a directory replication will fail. l With Mtree replication, resync will fail if the source MTree does not have retention lock enabled and the destination MTree has retention lock enabled. l With Mtree replication, resync will fail if the source and destination MTrees are retention lock enabled but the propagate retention lock option is set to FALSE.
418 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
Fastcopy When the filesys fastcopy [retention-lock] source src destination dest command is run on a system with a DD Retention Lock Governance enabled MTree, the command preserves the retention lock attribute during the fastcopy operation.
If the destination MTree is not retention lock enabled, the retention-lock file attribute is not preserved.
Filesys destroy Effects of the filesys destroy command when it is run on a system with a DD Retention Lock Governance enabled MTree.
l All data is destroyed, including retention-locked data. l All filesys options are returned to their defaults. This means that retention locking is disabled and the minimum and maximum retention periods are set back to their default values on the newly created file system.
This command is not allowed if DD Retention Lock Compliance is enabled on the system.
MTree delete When the mtree delete mtree-path command attempts to delete a DD Retention Lock Governance enabled (or previously enabled) MTree that currently contains data, the command returns an error.
The behavior of mtree delete is a similar to a command to delete a directory—an MTree with retention lock enabled (or previously enabled) can be deleted only if the MTree is empty.
DD Retention Lock compliance Certain DD OS commands behave differently when using DD Retention Lock Compliance. The following sections describe the differences for each.
Replication An MTree enabled with Retention Lock Compliance can be replicated via MTree and collection replication only. Directory replication is not supported. MTree and collection replication replicate the locked or unlocked state of files. Files that are compliance retention locked on the source are compliance retention locked on the destination and have the same level of protection. Minimum and maximum retention periods configured on MTrees are replicated to the destination system. To perform collection replication, the same security officer user must be present on both the source and destination systems before starting replication to the destination system and afterward for the lifetime of the source/replica pair.
DD Retention Lock compliance 419 DD Retention Lock
Replication Resync The replication resync destination command can be used with MTree replication, but not with collection replication.
l If the destination MTree contains retention-locked files that do not exist on the source, then resync will fail. l Both source and destination MTrees must be enabled for Retention Lock Compliance, or resync will fail.
Replication procedures The topics in this section describe MTree and collection replication procedures supported for Retention Lock Compliance.
For full descriptions of the commands referenced in the following topics, see the EMC Data Domain Operating System Command Reference Guide.
Replicating an MTree: One-to-one topology Replicate a Retention Lock Compliance enabled MTree from a source system to a destination system. Before you begin Enable DD Retention Lock on an MTree and configure client-side retention lock file control before replication. Procedure 1. Until instructed otherwise, perform the following steps on the destination system only. 2. Add the DD Retention Lock Compliance license on the system, if it is not present. a. First, check whether the license is already installed. license show b. If the RETENTION-LOCK-COMPLIANCE feature is not displayed, install the license. license add license-key
License keys are case-insensitive. Include the hyphens when typing keys.
3. Set up one or more security officer users accounts according to Role-Base Access Control (RBAC) rules. a. In the system administrator role, add a security officer account. user add user role security b. Enable the security officer authorization. authorization policy set security-officer enabled 4. Configure and enable the system to use DD Retention Lock Compliance.
420 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
Enabling DD Retention Lock Compliance enforces many restrictions on low-level access to system functions used during troubleshooting. Once enabled, the only way to disable DD Retention Lock Compliance is to initialize and reload the system, which results in destroying all data on the system.
a. Configure the system to use DD Retention Lock Compliance. system retention-lock compliance configure The system automatically reboots.
b. After the restart process is complete, enable DD Retention Lock Compliance on the system. system retention-lock compliance enable 5. Create a replication context. replication add source mtree://source-system-name/data/col1/ mtree-name destination mtree://destination-system-name/data/ col1/mtree-name 6. Perform the following steps on the source system only. 7. Create a replication context. replication add source mtree://source-system-name/data/col1/ mtree-name destination mtree://destination-system-name/data/ col1/mtree-name 8. Initialize the replication context. replication initialize mtree://destination-system-name/data/ col1/mtree-name 9. Confirm that replication is complete. replication status mtree://destination-system-name/data/ col1/mtree-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished. Replicating an MTree: One-to-many topology Replicate a Retention Lock Compliance enabled MTree from a source system to multiple destination systems. Before you begin Enable DD Retention Lock compliance on an MTree and configure client-side retention lock file control before replication. Procedure 1. Until instructed otherwise, perform the following steps on the destination system only. 2. Add the DD Retention Lock Compliance license on the system, if it is not present. a. First, check whether the license is already installed. license show b. If the RETENTION-LOCK-COMPLIANCE feature is not displayed, install the license. license add license-key
DD Retention Lock compliance 421 DD Retention Lock
License keys are case-insensitive. Include the hyphens when typing keys.
3. Set up one or more security officer users accounts according to Role-Base Access Control (RBAC) rules. a. In the system administrator role, add a security officer account. user add user role security b. Enable the security officer authorization. authorization policy set security-officer enabled 4. Configure and enable the system to use DD Retention Lock Compliance.
Enabling DD Retention Lock Compliance enforces many restrictions on low-level access to system functions used during troubleshooting. Once enabled, the only way to disable DD Retention Lock Compliance is to initialize and reload the system, which results in destroying all data on the system.
a. Configure the system to use DD Retention Lock Compliance. system retention-lock compliance configure The system automatically reboots.
b. After the restart process is complete, enable DD Retention Lock Compliance on the system. system retention-lock compliance enable 5. Create a replication context. replication add source mtree://source-system-name/data/col1/ mtree-name destination mtree://destination-system-name/data/ col1/mtree-name 6. Perform the following steps on the source system only. 7. Create a replication context for each destination system. replication add source mtree://source-system-name/data/col1/ mtree-name destination mtree://destination-system-name/data/ col1/mtree-name 8. Initialize the replication context for each destination system MTree. replication initialize mtree://destination-system-name/data/ col1/mtree-name 9. Confirm that replication is complete for each destination system. replication status mtree://destination-system-name/data/ col1/mtree-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished.
422 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
Adding Retention Lock Compliance protection to an existing MTree replication pair Add DD Retention Lock Compliance protection to an existing MTree replication pair that is not enabled for retention locking. Procedure 1. Until instructed otherwise, perform the following steps on both the source and destination systems. 2. Log in to the DD System Manager. The DD System Manager window appears with DD Network in the Navigation panel. 3. Select a Data Domain system. In the Navigation panel, expand DD Network and select a system 4. Add the Retention Lock Governance license, if it is not listed under Feature Licenses.
a. Select Administration > Licenses b. In the Licenses area click Add Licenses. c. In the License Key text box, type the license key.
License keys are case-insensitive. Include the hyphens when typing keys.
d. Click Add. 5. Break the current MTree context on the replication pair. replication break mtree://destination-system-name/data/col1/ mtree-name 6. Create the new replication context. replication add source mtree://source-system-name/data/col1/ mtree-name destination mtree://destination-system-name/data/ col1/mtree-name 7. Perform the following steps on the source system only. 8. Select an MTree for retention locking. Click the Data Management > MTree tab, then the checkbox for the MTree you want to use for retention locking. (You can also create an empty MTree and add files to it later.) 9. Click the MTree Summary tab to display information for the selected MTree. 10. Lock files in the compliance-enabled MTree. 11. Ensure that both source and destination (replica) MTrees are the same. replication resync mtree://destination-system-name/data/ col1/mtree-name 12. Check the progress of resync. replication watch mtree://destination-system-name/data/col1/ mtree-name 13. Confirm that replication is complete.
DD Retention Lock compliance 423 DD Retention Lock
replication status mtree://destination-system-name/data/ col1/mtree-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished. Converting a collection replication pair to MTree replication pairs A procedure for customers who used collection replication under DD Retention Lock Compliance in DD OS 5.2 and want to upgrade compliance-enabled MTrees in the collection replication pair to MTree replication pairs. Procedure 1. On the source system only: a. Create a snapshot for each Retention Lock Compliance enabled MTree. snapshot create snapshot-name /data/col1/mtree-name b. Synchronize the collection replication pair. replication sync col://destination-system-name c. Confirm that replication is complete. replication status col://destination-system-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished. d. View snapshot information for each Retention Lock Compliance enabled MTree. snapshot list mtree /data/col1/mtree-name Note the snapshot names for use later. 2. On the destination system only: a. Confirm that the replication is complete. replication status mtree://destination-system-name/data/ col1/mtree-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished. b. View each MTree snapshot replicated to the destination system. snapshot list mtree /data/col1/mtree-name c. Ensure that all Retention Lock Compliance MTree snapshots have been replicated by comparing the snapshot names generated here with those generated on the source system. snapshot list mtree /data/col1/mtree-name 3. On the both the source and destinations systems: a. Disable the file system. filesys disable b. Break the collection replication context. replication break col://destination-system-name c. Enable the file system. (Security officer authorization may be required.) filesys enable d. Add a replication context for each Retention Lock Compliance enabled MTree.
424 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
replication add source mtree://source-system-name/data/ col1/mtree-name destination mtree://destination-system- name/data/col1/mtree-name
Source and destination MTree names must be the same.
4. On the source system only: a. Ensure that both source and destination MTrees are the same. replication resync mtree://destination-system-name b. Check the progress of resync. replication watch destination c. Confirm that replication is complete. replication status mtree://destination-system-name/data/ col1/mtree-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished. Performing collection replication Replicate /data/col1 from a compliance-enabled source system to a compliance-enabled destination system.
For collection replication the same security officer account must be used on both the source and destination systems.
Procedure 1. Until instructed to do differently, perform the following steps on the source system only. 2. Log in to the DD System Manager. The DD System Manager window appears with DD Network in the Navigation Panel. 3. Select a Data Domain system. In the Navigation Panel, expand DD Network and select a system. 4. Add the Retention Lock Governance license, if it is not listed under Feature Licenses. a. Select Administration > Licenses b. In the Licenses area click Add Licenses. c. In the License Key text box, type the license key.
License keys are case-insensitive. Include the hyphens when typing keys.
d. Click Add. 5. Create the replication context. replication add source col://source-system-name destination col://destination-system-name
DD Retention Lock compliance 425 DD Retention Lock
6. Until instructed to do differently, perform the following steps on the destination system only. 7. Destroy the file system. filesys destroy 8. Log in to the DD System Manager. The DD System Manager window appears with DD Network in the Navigation Panel. 9. Select a Data Domain system. In the Navigation Panel, expand DD Network and select a system. 10. Create a file system, but do not enable it. filesys create 11. Create the replication context. replication add source col://source-system-name destination col://destination-system-name 12. Configure and enable the system to use DD Retention Lock Compliance. system retention-lock compliance configure (The system automatically reboots and executes the system retention-lock compliance enable command.) 13. Perform the following steps on the source system only. 14. Initialize the replication context. replication initialize source col://source-system-name destination col://destination-system-name 15. Confirm that replication is complete. replication status col://destination-system-name detailed This command reports 0 pre-compressed bytes remaining when replication is finished. Adding Retention Lock Compliance protection to an existing collection replication pair Add DD Retention Lock Compliance protection to a collection replication pair that was created without DD Retention Lock Compliance enabled on the source and destination systems. Procedure 1. Until instructed otherwise, perform the following steps on both the source and destination systems. 2. Disable the replication. replication disable col://destination-system-name 3. Log in to the DD System Manager. The DD System Manager window appears with DD Network in the Navigation Panel. 4. Select a Data Domain system. In the Navigation Panel, expand DD Network and select a system. 5. Until instructed otherwise, perform the following steps on the source system.
426 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
6. Configure and enable the system to use DD Retention Lock Compliance. system retention-lock compliance configure (The system automatically reboots by executing the system retention-lock compliance enable command.) 7. Enable the replication context. replication enable col://destination-system-name 8. Until instructed otherwise, perform the following steps on the destination system. 9. Configure and enable the system to use DD Retention Lock Compliance. system retention-lock compliance configure (The system automatically reboots by executing the system retention-lock compliance enable command.) 10. Enable the replication context. replication enable col://destination-system-name
Fastcopy When the filesys fastcopy [retention-lock] source src destination dest command is run on a system with a DD Retention Lock Compliance enabled MTree, the command preserves the retention lock attribute during the fastcopy operation.
If the destination MTree is not retention lock enabled, the retention-lock file attribute is not preserved.
CLI usage Considerations for a Data Domain system with DD Retention Lock Compliance.
l Commands that break compliance cannot be run. The following commands are disallowed:
n filesys archive unit del archive-unit
n filesys destroy
n mtree delete mtree-path
n mtree retention-lock disable mtree mtree-path
n mtree retention-lock revert
l The following command requires security officer authorization if the license being deleted is for DD Retention Lock Compliance:
n license del license-feature [license-feature . ] | license-code [license-code . ]
l The following commands require security officer authorization if DD Retention Lock Compliance is enabled on an MTree specified in the command:
DD Retention Lock compliance 427 DD Retention Lock
n mtree rename mtree-path new-mtree-path
l The following commands require security officer authorization if DD Retention Lock Compliance is enabled on the system:
n alerts notify-list reset
n config set timezone zonename
n config reset timezone
n cifs set authentication active-directory realm < [dc1 [dc2 . ]]
n license reset
n ntp add timeserver time server list
n ntp del timeserver time server list
n ntp reset timeservers
n replication break
n replication disable
n system set date MMDDhhmm[[CC]YY]
System clock DD Retention Lock Compliance implements an internal security clock to prevent malicious tampering with the system clock. The security clock closely monitors and records the system clock. If there is an accumulated two-week skew within a year between the security clock and the system clock, the Data Domain file system (DDFS) is disabled and can be resumed only by a security officer. Finding the System Clock Skew You can run the DD OS command system retention-lock compliance status (security officer authorization required) to get system and security clock information, including the last recorded security clock value, and the accumulated system clock variance. This value is updated every 10 minutes. Removing the system clock skew Clock skew is updated every time the security clock records a new value for the system clock. After 1 year, it is reset to 0. At any time, you can run the DD OS command system set date MMDDhhmm[[CC]YY] to set the time of the system clock (security officer authorization required). If the clock skew becomes larger than the preset value (2 weeks), the file system is disabled. Complete these steps to restart the file system and remove the skew between security and system clocks. Procedure 1. At the system console, enable the file system. filesys enable 2. At the prompt, confirm that you want to quit the filesys enable command and check whether the system date is right.
428 EMC Data Domain Operating System 5.7 Administration Guide DD Retention Lock
3. Display the system date. system show date 4. If the system date is not correct, set the correct date (security officer authorization is required) and confirm it. system set date MMDDhhmm[[CC]YY] system show date 5. Enable the file system again. filesys enable 6. At the prompt, continue to the enabling procedure. 7. A security officer prompt appears. Complete the security officer authorization to start the file system. The security clock will automatically be updated to the current system date.
DD Retention Lock compliance 429 DD Retention Lock
430 EMC Data Domain Operating System 5.7 Administration Guide CHAPTER 18 DD Encryption
This chapter includes:
l DD encryption overview. 432 l Configuring encryption. 433 l About key management. 434 l Key manager setup. 438 l Changing key managers after setup. 441 l Checking settings for encryption of data at rest. 443 l Enabling and disabling encryption of data at rest. 443 l Locking and unlocking the file system. 444
DD Encryption 431 DD Encryption
DD encryption overview Data encryption protects user data if the Data Domain system is stolen or if the physical storage media is lost during transit, and it eliminates accidental exposure of a failed drive if it is replaced. When data enters the Data Domain system using any of the supported protocols (NFS, CIFS, VTL, DD Boost, and NDMP Tape Server), the stream is segmented, fingerprinted, and de-duplicated (global compression). It is then grouped into multi-segment compression regions, locally compressed, and encrypted before being stored to disk. Once enabled, the Encryption at Rest feature encrypts all data entering the Data Domain system. You cannot enable encryption at a more granular level.
Data that has been stored before the encryption feature is enabled does not automatically get encrypted. To protect all of the data on the system, be sure to enable the option to encrypt existing data when you configure encryption.
Additional Notes: As of DD OS 5.5.1.0, Encryption of Data at Rest is supported for DD Extended Retention- enabled systems with a single retention unit. As of 5.5.1.0, DD Extended Retention supports only a single retention unit, so systems set up during, or after, 5.5.1.0 will have no problem complying with this restriction. However, systems set up prior to 5.5.1.0 may have more than one retention unit, but they will not work with Encryption of Data at Rest until all but one retention unit has been removed, or data has been moved or migrated to one retention unit. The filesys encryption apply-changes command applies any encryption configuration changes to all data present in the file system during the next cleaning cycle. For more information about this command, see the EMC Data Domain Operating System Command Reference Guide. Encryption of Data at Rest supports all of the currently supported backup applications described in the Backup Compatibility Guides available through EMC Online Support at http://support.emc.com. Data Domain Replicator software can be used with the encryption option, enabling encrypted data to be replicated using collection, directory, MTree, or application-specific managed file replication with the various topologies. Each replication form works uniquely with encryption and offers the same level of security. For more information, see the section on using encryption of data at rest with replication. Files locked using the Data Domain Retention Lock software options can be stored, encrypted, and replicated. The autosupport feature includes information about the state of encryption on the Data Domain system:
l Whether or not encryption is enabled l The Key Manager in effect and which keys are used l The encryption algorithm that is configured l The state of the file system
432 EMC Data Domain Operating System 5.7 Administration Guide DD Encryption
Configuring encryption This procedure includes configuring a key manager. If the Encryption Status on the Data Management > File System > Encryption tab shows Not Configured, click Configure to set up encryption on the Data Domain system. Provide the following information: l Algorithm n Select an encryption algorithm from the drop-down list or accept the default AES 256-bit (CBC). The AES 256-bit Galois/Counter Mode (GCM) is the most secure algorithm but it is significantly slower than the Cipher Block Chaining (CBC) mode. n Determine what data is to be encrypted: existing and new or only new. Existing data will be encrypted during the first cleaning cycle after the file system is restarted. Encryption of existing data can take longer than a standard file system cleaning operation. l Key Manager (select one of the two) n Embedded Key Manager By default, the Data Domain Embedded Key Manager is in effect after you restart the file system unless you configure the RSA DPM Key Manager. You can enable or disable key rotation. If enabled, enter a rotation interval between 1-12 months. n RSA DPM Key Manager
See the section about key management for an explanation about how the Embedded Key Manager and the RSA DPM Key Manager work.
The RSA DPM Key Manager requires setup on both an RSA DPM server and on the Data Domain system. Follow the instructions in the RSA DPM key manager encryption setup section before selecting the RSA DPM Key Manager in the Data Domain interface. You can enable encryption using the Embedded Key Manager before configuring the RSA DPM Key Manager. You can then switch to the RSA DPM Key Manager after performing an RSA DPM key manager encryption setup and following the procedure described in the changing key managers after setup section.
Table 173 Configuring RSA Key Manager
l Enter the name or the IP address of the Key Manager server.
l Choose the key class that the Data Domain system will use to generate the key from the menu.
l Enter the port number (443 is the default).
l Select whether the imported host certificate is FIPS compliant. The default mode is enabled.
Configuring encryption 433 DD Encryption
The Summary shows your selected configuration values. Review them for correctness. To change a value, click Back to navigate to the page where it was entered and modify it. A system restart is necessary to enable encryption. To apply the new configuration, select the option to restart the file system.
Applications may experience an interruption while the file system is restarted.
About key management Encryption keys determine the output of the cryptographic algorithm. They are protected by a passphrase, which encrypts the encryption key before it is stored in multiple locations on disk. The passphrase is user generated and requires both an administrator and a security officer to change it. A key manager controls the generation, distribution, and lifecycle management of multiple encryption keys. A Data Domain system can use either the Embedded Key Manager or the RSA Data Protection Manager (DPM) Key Manager. Only one can be in effect at a time. When encryption is enabled on a Data Domain system, the Embedded Key Manager is in effect by default. If you configure the RSA DPM Key Manager, it replaces the Embedded Key Manager and remains in effect until you disable it. A file system restart is required for a new key manager to be operational. Both key managers provide multiple keys, although the system uses only one key at a time to encrypt data coming into a Data Domain system. If the RSA DPM Key Manager is configured and enabled, the Data Domain systems use keys provided by the RSA DPM Key Manager Server. If the same DPM Key Manager manages multiple Data Domain systems, all systems will have the same active key (if they are using the same key class) when the systems are synced and the Data Domain file system has been restarted. The Embedded Key Manager generates its keys internally. Both key managers rotate keys and support a maximum of 254 keys. The Embedded Key Manager allows you to specify how many months a key is in effect before being replaced (after the file system is restarted). The RSA DPM Key Manager rotates keys on a regular basis, depending on the key class. The Embedded Key Manager key rotation is managed on the Data Domain system. The DPM Key Manager key rotation is managed on the RSA DPM Key Manager server. The section covers the following major topics.
Rectifying lost or corrupted keys Create a file that contains all of your system’s current encryption keys. Your support provider can use this file to import keys back to your system should they become lost or corrupted. It is recommended that you create an export file on a regular basis. You are prompted for the Security Officer's credential to export the keys. For additional key file protection, you can use a passphrase that differs from the one used in a Data Domain system. After exporting, it is recommended that you save the key file in a secure file server accessible only by authorized users. You must remember the passphrase used for the key file. If the passphrase is lost or forgotten, the Data Domain system cannot import and restore the keys. Enter: # filesys encryption keys export
434 EMC Data Domain Operating System 5.7 Administration Guide DD Encryption
Key manager support Both Key Managers support all DD OS file system protocols. Replication When configuring Data Domain systems for directory MTree replication, configure each Data Domain system separately. The two systems can use either the same or a different key class, and the same or different key managers. For collection replication configuration, the Data Domain system must be configured on the source. After a replication break, the original replica Data Domain system has to be configured for the Key Manager. If not, the Data Domain system continues to use the latest known key.
Working with the RSA DPM Key Manager If the RSA DPM Key Manager is configured and enabled, the Data Domain systems use keys provided by the RSA DPM Key Manager Server. If the same DPM Key Manager manages multiple Data Domain systems, all systems will have the same active key (if they are using the same key class) when the systems are synced and the Data Domain file system has been restarted. Key rotation is managed on the RSA DPM Key Manager server. If the RSA DPM Key Manager is configured and enabled, the Data Domain systems use keys provided by the RSA DPM Key Manager Server. If the same DPM Key Manager manages multiple Data Domain systems, all systems will have the same active key (if they are using the same key class) when the systems are synced and the Data Domain file system has been restarted. Key rotation is managed on the RSA DPM Key Manager server. Encryption Key States One Activated-RW key is always in effect. If the active key is compromised, the RSA DPM Key Manager provides a new key. When the Data Domain system detects the new key, it issues an alert for the administrator to restart the file system. Expired keys become read only for the existing data on the Data Domain system, and a new active key is applied to all new data that is ingested. When a key is compromised, the existing data is re-encrypted using the new encryption key after a file system cleaning is run. If the maximum number of keys is reached, unused keys must be deleted to make room for new keys. To view information about the encryption keys that are on Data Domain system, open the DD System Manager and go to the Data Management > File System > Encryption tab. Keys are listed by ID number in the Encryption Keys section of the Encryption tab. The following information is given for each key: when a key was created, how long it is valid, its type (RSA DPM or Data Domain), its state (see DPM Encryption Key States Supported by Data Domain), and its post-compression size. If the system is licensed for Extended Retention, the following fields are also displayed: Active Size (post comp) The amount of physical space on the active tier encrypted with the key. Retention Size (post comp) The amount of physical space on the retention tier encrypted with the key.
Click on a Key MUID and the system displays the following information for the key in the Key Details dialog: Tier/Unit (example: Active, Retention-unit-2), creation date, valid until date, state (see DPM Encryption Key States Supported by Data Domain), and post compression size. Click Close to close the dialog.
Key manager support 435 DD Encryption
Table 174 DPM encryption key states supported by Data Domain
Pending-Activated The key has just been created. After a file system restart, the key becomes Activated-RW.
Activated-RW and Both Activated-RW and Activated-RO read the data encrypted with their Activated-RO keys, respectively. Activated-RW is the latest activated key.
De-Activated A key becomes deactivated when the current time exceeds the validity period. The key is used for reading.
Compromised The key can only decrypt. After all of the data encrypted with the compromised key is re-encrypted, the state changes to Destroyed Compromised. The keys are re-encrypted when a file system cleaning is run. You can delete a Destroyed Compromised key, if necessary.
Marked-For-Destroy You have marked the key as destroyed for the data to be re-encrypted.
Destroyed After re-encrypting all data encrypted with this key, the DD OS changes it from Marked-For-Destroy to Destroyed. Also, when the key that is destroyed is compromised, it becomes Compromised-Destroyed. You can delete keys that are Destroyed and Compromised-Destroyed.
A key is not destroyed in the Data Domain system until a cleaning operation is run and completed.
Keeping keys in sync with the RSA DPM key manager An automatic key sync is performed every day at midnight. A manual key sync is required only if you cannot wait for the scheduled sync. Whenever new keys are synced on the Data Domain system, an alert is generated. This alert is cleared after the Data Domain file system is restarted. After the RSA DPM Key Manager Server generates new keys, click the Sync button to have them display in the Encryption Key list on the Data Domain System Manager’s Encryption tab.
A file system restart is necessary if keys have changed since the last sync.
Procedure 1. Using the DD System Manager, select the Data Domain system you are working with in the Navigation panel.
Always perform DD System Manager functions on the system you have selected in the Navigation panel.
2. Select Data Management > File System > Encryption. 3. In the Encryption Keys section, select the RSA DPM key, and click Sync.
436 EMC Data Domain Operating System 5.7 Administration Guide DD Encryption
Destroying a key (RSA DPM key manager) Destroy a key if you do not want any data to be encrypted with it. This procedure requires security officer credentials.
For information about the security officer, see the sections regarding creating local users and enabling security authorization.
To change an RSA DPM key to a state in which it can be destroyed: Procedure 1. Deactivate the key on the RSA DPM Server. 2. Restart the file system for the key to be deactivated on the Data Domain system. 3. Using the DD System Manager, select Data Management > File System > Encryption. 4. In the Encryption Keys section, select the key in the list to be destroyed. 5. Click Destroy. The system displays the Destroy dialog that includes the tier and state for the key. 6. Enter your security officer user name and password. 7. Confirm that you want to destroy the key by clicking Destroy.
After a file system clean has run, the key state will change to Destroyed.
Deleting a key You can delete Key Manager keys that are in the Destroyed or Compromised-Destroyed states. However, you only need to delete a key when the number of keys has reached the maximum 254 limit. This procedure requires security officer credentials.
To reach the Destroyed state, the Destroying a Key procedure (for either the Embedded Key Manager or the RSA DPM Key Manager) must be performed on the key and a system cleaning must be run.
Procedure 1. Select Data Management > File System > Encryption. 2. In the Encryption Keys section, select the key or keys in the list to be deleted. 3. Click Delete. The system displays the key to be deleted, and the tier and state for the key. 4. Type your security officer user name and password. 5. Confirm that you want to delete the key or keys by clicking Delete.
Working with the Embedded Key Manager When the Embedded Key Manager is selected, the Data Domain System creates its own keys. After the key rotation policy is configured, a new key is automatically created at the next rotation. An alert informs you of the creation of a new key. You must perform a file system
Working with the Embedded Key Manager 437 DD Encryption
restart to activate the new key and deactivate the old key. You can disable the key rotation policy by clicking the disable button associated with the Embedded Key Manager Key’s rotation status.
Creating a key (Embedded Key Manager) Create an encryption key for the Embedded Key Manager. Procedure
1. Select Data Management > File System > Encryption. 2. In the Encryption Keys section, click Create. . 3. Type your security officer user name and password. 4. Click Restart the filesystem now if you want to restart the file system. A new Data Domain key will be created. After the file system is restarted, the previous key will become deactivated and the new key will become activated. 5. Click Create.
Destroying a key (Embedded Key Manager) Destroy an encryption key for the Embedded Key Manager. Procedure
1. Select Data Management > File System > Encryption. 2. In the Encryption Keys section, select the key in the list to be destroyed. 3. Click Destroy. The system displays the Destroy dialog that includes the tier and state for the key. 4. Type your security officer user name and password. 5. Confirm that you want to destroy the key by clicking Destroy.
After a file system clean has run, the key state changes to Destroyed.
How the cleaning operation works Encryption affects the performance of cleaning operations when data encrypted with the Compromised or Marked-For-Destroyed keys is re-keyed using the Activated-RW key. At the end of the cleaning operation, there will be no data that is encrypted with the Compromised or Marked-For-Destroyed keys. Also, any data written by the cleaning operation is encrypted with the Activated-RW key.
Key manager setup Follow the instructions for the type of key manager you are using.
RSA DPM Key Manager encryption setup RSA DPM Key Manager must be set up on both the RSA DPM Server and on the Data Domain system.
438 EMC Data Domain Operating System 5.7 Administration Guide DD Encryption
Performing this setup on the RSA DPM server The main steps for setting up the RSA DPM Server (using its graphical user interface).
See the latest version of the RSA Data Protection Manager Server Administrator’s Guide for more information about each step of this procedure. Algorithm and cipher mode settings set on the RSA DPM Key Manager Server are ignored by the Data Domain system. Configure these settings on the Data Domain system.
Procedure 1. Create an identity for the Data Domain system using the X509 certificate. A secure channel is created based on this certificate. 2. Create a key class with the proper attributes: l Key length: 256 bits. l Duration: For example, six months or whatever matches your policy. l Auto-key generation: Select to have keys automatically generated.
Multiple Data Domain systems can share the same key class. For more information about key classes, see the section about RSA DPM key classes.
3. Create an identity using the Data Domain system’s host certificate as its identity certificate. The identity and the key class have to be in the same identity group. 4. Import the certificates. See the section about importing certificates for more information.
About RSA DPM key classes The Data Domain system retrieves a key from RSA DPM Key Manager by key class. A key class is a specialized type of security class used by the RSA DPM Key Manager that groups crytopgraphic keys with similar characteristics. The RSA DPM Key Manager Server allows a key class to be set up to either return the current key, or to generate a new key each time. The Data Domain system supports only the key classes configured to return the current key. Do not use a key class that is configured to generate a new key each time.
If the key length is not 256 bits, the DPM configuration will fail.
Importing the certificates After obtaining certificates, import them to the Data Domain system. Before you begin l The Host certificate should be in PKCS12 format. l The CA certificate should be in PEM format. l You must obtain CA and Host certificates that are compatible with the RSA DPM Key Manager. You can request these certificates from third-party certificates authorities, or create them using appropriate SSL utility tools.
RSA DPM Key Manager encryption setup 439 DD Encryption
l If the system passphrase is not set, you cannot import the host certificate. The passphrase is set when you enable encryption. To change it, see the section regarding changing the system passphrase in the Managing the Data Domain Systems chapter. DD OS supports certificates without any extension and certificates with server and client extensions for use with both the Data DD Manager and RSA DPM Key Manager. Certificates with client extensions are supported only by RSA DPM Key Manager, and certificates with server extensions are supported only by the DD System Manager. DD OS does not support the RSA DPM Key Manager Server’s Auto Registration Certificate capability, which uploads an auto registered certificate directly, or imports multiple certificates. This means that you must import the CA and Host certificates for a Data Domain system. The following information describes how to respond to a couple of alerts that might appear during certificate management.
l If HTTPS fails to restart due to corrupted imported certificates, self-signed certificates are used. If this occurs, a managed alert, UnusableHostCertificate, is issued. To clear the alert, delete the corrupted certificates and re-import new certificates. l If imported certificates are removed, for example during a system headswap and the imported certificates fail to copy over, a managed alert, MissingHostCertificate, is issued. Re-import the certificates to clear the alert. After obtaining the certificates, import them to the Data Domain system as follows: Procedure 1. Configure the RSA DPM Key Manager Server to use the CA and Host certificates. For instructions, see the RSA DPM Key Manager Server Administration Guide. 2. Import the certificates by redirecting the certificate files using ssh command syntax. See the EMC Data Domain Operating System Command Reference Guide for details.
ssh sysadmin@ adminaccess certificate import File System > Encryption tab. 5. Follow the instructions in the section regarding configuring encryption and select the DPM Key Manager. If encryption has already been set up, follow the instructions in the section regarding changing key managers after setup.
Changing key managers after setup Select from the Embedded Key Manager or the RSA DPM Key Manager. Before you begin To manage certificates for a system, you must start DD System Manager on that system. Procedure
1. Select Data Management > File System > Encryption. 2. Under Key Management, click Configure. 3. Type your security officer username and password. 4. Select which Key Manager to use.
l Embedded Key Manager: Select to enable or disable key rotation. If enabled, enter a rotation interval between 1-to-12 months. Select Restart the file system now, and click OK. l RSA DPM Key Manager: Enter the server name, key class, port (the default is 443), and whether the imported host certificate is FIPS compliant. The default mode is enabled. Select Restart the file system now, and click OK.
5. Click Manage Certificates to add certificates.
Changing key managers after setup 441 DD Encryption
Managing certificates for RSA Key Manager You must use both host and CA certificates with RSA Key Manager.
Certificates are only necessary for RSA Key Manager. Embedded Key Manager does not use certificates.
Adding CA certificates for RSA Key Manager Upload or copy and paste CA certificates. Procedure 1. Select one of the following:
l Select the option to upload a CA certificate as a .pem file and click Browse to find the file. l Select the option to copy and paste the CA certificate and paste the certificate contents into the field provided. 2. Click Add to add the certificate.
Adding a host certificate for RSA Key Manager Upload the certificate as a .p12 file, or upload a public key as a .pem file and use a generated private key. To start, select the first or second of the following steps: Procedure 1. Select the option to upload the certificate as a .p12 file. a. Enter a password. b. Click Browse to find the .p12 file. 2. Select the option to upload the public key as a .pem file and use a generated private key. a. Click Browse to find the .pem file. 3. Click Add.
Deleting certificates Select a certificate with the correct fingerprint. Procedure 1. Select a certificate to delete. 2. Click Delete. The system displays a Delete Certificate dialog with the fingerprint of the certificate to be deleted. 3. Click OK.
442 EMC Data Domain Operating System 5.7 Administration Guide DD Encryption
Checking settings for encryption of data at rest Check the settings for the Encryption feature. Click the Data Management > File System > Encryption tabs. The currently used Key Manager is shown as Enabled. For a description of the Encryption settings, see the section about the encryption view.
Enabling and disabling encryption of data at rest After configuring Encryption, the status is enabled and the Disabled button is active. When Encryption is disabled, the Enabled button is active.
Enabling encryption of data at rest Use the DD System Manager to enable the Encryption feature. Procedure 1. Using the DD System Manager, select the Data Domain system you are working with in the Navigation panel. 2. In the Encryption view, click the Enable button. 3. Both of the following options are available:
l Select Apply to existing data and click OK. Encryption of existing data will occur during the first cleaning cycle after the file system is restarted.
l Select Restart the file system now and click OK. Encryption will be enabled after the file system is restarted. After you finish
Applications may experience an interruption while the file system is restarted.
Disabling encryption of data at rest Use the DD System Manager to disable the Encryption feature. Procedure 1. Using the DD System Manager, select the Data Domain system you are working with in the Navigation panel. 2. In the Encryption view, click the Disable button. The Disable Encryption dialog box is displayed.
3. In the Security Officer Credentials area, enter the user name and password of a security officer. 4. Select one of the following:
l Select Apply to existing data and click OK. Decryption of existing data will occur during the first cleaning cycle after the file system is restarted.
l Select Restart the file system now and click OK. Encryption will be disabled after the file system is restarted.
Checking settings for encryption of data at rest 443 DD Encryption
After you finish
Applications may experience an interruption while the file system is restarted.
Locking and unlocking the file system Use this procedure when an encryption-enabled Data Domain system (and its external storage devices) are being transported, or if you want to lock a disk that is being replaced. The procedure requires two accounts: Security Officer and System Administration roles. Procedure
1. Select Data Management > File System > Encryption . In the File System Lock area, the Status shows whether the file system is Locked or Unlocked.
2. Disable the file system by clicking Disabled in the File System status area. 3. Use the procedure to lock or unlock the file system.
Locking the file system To lock the file system, Encryption must be enabled and the file system must be disabled. Procedure
1. Select Data Management > File System > Encryption and click Lock File System. 2. In the text fields of the Lock File System dialog box, provide:
l The username and password of a Security Officer account (an authorized user in the Security User group on that Data Domain system). l The current and a new passphrase.
3. Click OK. This procedure re-encrypts the encryption keys with the new passphrase. This process destroys the cached copy of the current passphrase (both in-memory and on-disk).
Changing the passphrase requires two-user authentication to protect against the possibility of a rogue employee‘s shredding the data.
Be sure to take care of the passphrase. If the passphrase is lost, you will never be able to unlock the file system and access the data. The data will be irrevocably lost.
4. Shut down the system:
444 EMC Data Domain Operating System 5.7 Administration Guide DD Encryption
Do not use the chassis power switch to power off the system. Type the following command at the command prompt instead.
# system poweroff The ‘system poweroff’ command shuts down the system and turns off the power. Continue? (yes|no|?) [no]: 5. Transport the system or remove the disk being replaced. 6. Power on the system and use the procedure to unlock the file system.
Unlocking the file system This procedure prepares an encrypted file system for use after it has arrived at its destination. Procedure
1. Select Data Management > File System > Encryption and click Unlock File System. 2. In the text fields, type the passphrase that was used to lock the file system. 3. Click OK. 4. Click Close to exit. If the passphrase is incorrect, the file system does not start and the system reports the error. Type the correct passphrase, as directed in the previous step.
Changing the encryption algorithm Reset the encryption algorithm if necessary, or select options to encrypt new and existing data or just new data. Procedure
1. Select Data Management > File System > Encryption 2. To change the Encryption Algorithm used to encrypt the Data Domain system, click Change Algorithm. The Change Algorithm dialog box is displayed. Supported encryption algorithms are:
l AES-128 CBC l AES-256 CBC l AES-128 GCM l AES-256 GCM
3. Select an encryption algorithm from the drop-down list or accept the default AES 256- bit (CBC). The AES 256-bit Galois/Counter Mode (GCM) is the most secure algorithm but it is significantly slower than the Cipher Block Chaining (CBC) mode.
To reset the algorithm to the default AES 256-bit (CBC), click Reset to default.
4. Determine what data will be encrypted:
Unlocking the file system 445 DD Encryption
l To encrypt existing and new data on the system, select Apply to Existing data, Restart file system now, and click OK. Existing data will be encrypted during the first cleaning cycle after the file system is restarted.
Encryption of existing data can take longer than a standard file system clean operation.
l To encrypt only new data, select Restart file system now and click OK.
5. The status is displayed. Click Close when the process is complete.
Applications may experience an interruption while the file system is restarted.
446 EMC Data Domain Operating System 5.7 Administration Guide