EMC NetWorker Module for Microsoft Applications

Transcription

EMC® NetWorker®
Module for Microsoft Applications
Release 2.2
Administration Guide
P/N 300-008-851
REV A02
EMC Corporation
Corporate Headquarters:
Hopkinton, MA 01748-9103
1-508-435-1000
www.EMC.com
Copyright © 2007-2009 EMC Corporation. All rights reserved.
Published July, 2009
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.
For the most up-to-date regulatory document for your product line, go to the Document/Whitepaper Library on EMC
Powerlink.
For the most up-to-date listing of EMC product names, see EMC Corporation Trademarks on EMC.com.
All other trademarks used herein are the property of their respective owners.
2
EMC NetWorker Module for Microsoft Applications Release 2.2 Administration Guide
Contents
Preface
Chapter 1
Overview
NetWorker Module for Microsoft Applications .................................................
NetWorker Module features ...........................................................................
Privileged user-level access.............................................................................
Windows and Microsoft application updates ..............................................
Supported configurations......................................................................................
Supported NetWorker Server hosts...............................................................
Windows operating system support .............................................................
Simple LAN-based configuration ..................................................................
LAN-based configuration with a proxy client .............................................
LAN-free configuration ...................................................................................
The VSS snapshot creation process ......................................................................
Writers................................................................................................................
Providers............................................................................................................
Dynamic volume support ...............................................................................
Providers and backup performance ..............................................................
NMM client snapshot backups .............................................................................
Nonpersistent rollover backup.......................................................................
Instant backup with or without rollover.......................................................
Serverless backup .............................................................................................
NMM client recoveries...........................................................................................
Conventional recovery.....................................................................................
Instant recovery ................................................................................................
Rollback recovery .............................................................................................
Disaster and granular backup and recovery for applications....................
The NMM client interface......................................................................................
Recover view .....................................................................................................
Snapshot Management view...........................................................................
Monitor view.....................................................................................................
NMM client graphical user interface ...................................................................
Opening the software and connecting to a NetWorker server ..................
Selecting an item for recovery ........................................................................
Searching for an item .......................................................................................
Specifying a recovery browse time ................................................................
Selecting a backup version for recovery .......................................................
Viewing the volumes required for a recovery..............................................
20
20
21
21
22
22
22
23
24
25
27
28
30
30
31
32
32
32
32
34
34
34
34
35
37
37
39
40
41
41
42
42
43
43
44
3
Contents
Chapter 2
Configuring a Scheduled Backup
Backup configuration roadmap ............................................................................ 46
Privileged user-level access ............................................................................. 47
Name resolution................................................................................................ 48
Best practices and considerations for backing up application data........... 48
Task 1: Configure a backup pool for snapshot operations ................................ 50
Configuring a device ........................................................................................ 50
Configuring a label template........................................................................... 51
Configuring a backup pool.............................................................................. 51
Labeling the device ........................................................................................... 52
Task 2: Configure snapshot policies ..................................................................... 53
Snapshot policies for SYSTEM COMPONENTS backup ............................ 53
Creating a snapshot policy .............................................................................. 53
Snapshot policy examples................................................................................ 55
Task 3: Configure a backup schedule ................................................................... 56
Task 4: Configure a backup group ........................................................................ 57
Task 5: Configure a Client resource ...................................................................... 59
Configuring a Client resource for System volumes, components,
and applications ................................................................................................ 59
Specifying save sets for application data....................................................... 59
Application information variable settings .................................................... 63
Task 6: Configure privileges .................................................................................. 64
Task 7: Configure a proxy client............................................................................ 65
Backing up a clustered NMM Client .................................................................... 67
NetWorker pathownerignore cluster functionality is not supported ....... 67
Cluster failover and backups........................................................................... 67
Moving an NMM Client to another NetWorker server ..................................... 68
Chapter 3
Deduplication Backup and Recovery
Data deduplication.................................................................................................. 70
Related information about Avamar server and NetWorker server........... 71
Avamar integration in NMM .......................................................................... 71
Windows OS, application, and configuration support ............................... 72
Unsupported for deduplication ...................................................................... 72
Avamar and NetWorker server configuration .................................................... 73
Configuring Avamar and the NetWorker server to receive
deduplication data and process backups ...................................................... 73
Setting DNS resolution for NetWorker Deduplication Node
and NMM .......................................................................................................... 73
Deduplication backups........................................................................................... 74
Backup levels and schedules for deduplication clients............................... 74
Retention policies .............................................................................................. 74
Backup configuration notes............................................................................. 75
Configuring a Client Resource for deduplication .............................................. 76
Configuring a deduplication client ................................................................ 76
Recovering deduplicated data .............................................................................. 78
Deleting deduplication save sets .................................................................... 78
Querying deduplication save sets using mminfo ........................................ 78
Additional NetWorker server deduplication operation information ....... 78
Chapter 4
System Backup and Recovery
Windows system backup ....................................................................................... 80
Windows system support ................................................................................ 80
4
Contents
Specifying system save sets for recovery ...................................................... 80
NMM System data commands ....................................................................... 81
Performing system backups .................................................................................. 82
Task 5: Configure a System Client resource ................................................. 82
Performing a recovery............................................................................................ 84
System recovery in Windows Server 2008 and Windows Server 2003..... 85
System Recovery Options Summary ............................................................ 86
Recovery options .............................................................................................. 87
General recovery options ................................................................................ 87
NetWorker recovery options .......................................................................... 87
Cluster recovery options ................................................................................. 89
ADAM recovery options ................................................................................. 91
DFS recovery options ....................................................................................... 92
FRS recovery options ....................................................................................... 95
NTDS recovery options ................................................................................... 96
Security recovery options................................................................................ 98
Recovering file system snapshots that have not been rolled over................... 99
Booting Windows in Directory Service Restore mode .................................... 100
Performing an authoritative recovery of NTDS or FRS data ......................... 101
Performing a directed recovery .......................................................................... 103
Peforming a directed recovery with NMM ................................................ 103
Recovering the Windows system configuration to an earlier state ............... 106
Disaster recovery .................................................................................................. 107
Disaster recovery when the NMM client is not on a domain
controller.......................................................................................................... 108
Disaster recovery when the NMM client is on a domain
controller.......................................................................................................... 109
Chapter 5
Microsoft SQL Server Backup and Recovery
Microsoft SQL Server backup and recovery .....................................................
SQL Server versions supported by NMM...................................................
Types of supported backup and recovery ..................................................
Components used by NMM for SQL data backup and recovery............
Specifying save sets for SQL data ................................................................
Displaying valid SQL data save sets............................................................
URL encoding for SQL save sets ..................................................................
SQL Server application information variable settings ..............................
Serverless backup ...........................................................................................
Performing SQL Server backups ........................................................................
Full backup of an SQL Server .......................................................................
Performing SQL Server recovery........................................................................
Recovering SQL Server 2005 and 2008 stand-alone databases
from the SQLServerWriter save set..............................................................
Recovering SQL Server 2005 or 2008 express databases from
the System Components save set .................................................................
Best practices for SQL Server backup and recovery..................................
Performing SQL Server disaster recovery.........................................................
SQL Server (standalone) disaster recovery on Windows Server 2008....
SQL Server (standalone) disaster recovery on Windows Server 2003....
SQL Server cluster disaster recovery on Windows Server 2008 ..............
SQL Server cluster disaster recovery on Windows Server 2003 ..............
112
112
112
113
113
114
115
116
116
117
117
119
119
119
119
121
121
121
122
123
5
Contents
Chapter 6
Microsoft SharePoint Server Backup and Recovery
Microsoft SharePoint Server backup and recovery .......................................... 126
SharePoint Server versions supported by NMM client............................. 126
SharePoint Server 2007 backup and recovery ................................................... 127
Types of SharePoint Server 2007 backup and recovery ............................ 127
Protecting a SharePoint farm......................................................................... 128
Storage requirements and performance impact in SharePoint
backup .............................................................................................................. 129
NMM installation requirements for a Microsoft Office SharePoint
2007 system ...................................................................................................... 129
SharePoint 2007 data backed up by NMM.................................................. 131
Keeping backups and recoveries in sync..................................................... 131
Components used by NMM for SharePoint backup and recovery ......... 132
Specifying SharePoint 2007 save sets for application data ....................... 133
Displaying valid SharePoint 2007 data save sets ....................................... 133
URL encoding for SharePoint save sets ....................................................... 134
SharePoint 2007 application information variable settings ...................... 135
Performing SharePoint 2007 backups ................................................................ 136
Configuring SharePoint 2007 backups......................................................... 136
Task 5: Configure a SharePoint 2007 Client resource ....................................... 138
Performing a full backup for a distributed SharePoint farm.................... 143
Performing SharePoint 2007 recovery................................................................ 143
Mandatory order for SharePoint recoveries................................................ 144
Configuration database recovery requirements in full farm
recovery ............................................................................................................ 144
Selecting valid items for recovery ................................................................ 144
Full recovery of a stand-alone SharePoint farm ......................................... 144
Full recovery of a distributed SharePoint farm .......................................... 145
Recovery of individual SharePoint content databases .............................. 146
Recovery of SharePoint Search/Content Indexes ...................................... 146
Rollback of SharePoint SQL databases ........................................................ 147
SharePoint 2003 backup and recovery ............................................................... 147
Types of supported backup and recovery................................................... 147
Components used by NMM for SQL backup and recovery ..................... 148
Specifying save sets for SharePoint 2003 SQL data.................................... 148
Displaying valid SQL data save sets for SharePoint 2003 backup........... 149
URL encoding for SQL save sets in SharePoint 2003 backup................... 150
SQL application information variable settings in SharePoint
2003 backup ..................................................................................................... 151
Performing SharePoint 2003 backups ................................................................ 152
Task 5: Configuring a SharePoint 2003 Client resource ............................ 152
Performing a SharePoint 2003 recovery............................................................. 154
Recovering SQL Server 2000 stand-alone databases for
SharePoint 2003 ............................................................................................... 154
Recovering SQL 2005 databases for SharePoint 2003 ................................ 154
Best practices for SQL Server backup and recovery of
SharePoint 2003 databases ............................................................................. 156
SharePoint 2003 scenario of backup and recovery process ............................. 158
SharePoint 2007 disaster recovery ...................................................................... 165
Disaster recovery planning............................................................................ 165
Performing a disaster recovery ..................................................................... 167
6
Contents
Chapter 7
Microsoft SharePoint 2007 Granular Backup and Recovery
NMM support for granular backup and recovery of SharePoint items .......
SharePoint objects available in NMM granular backup ...........................
SharePoint limitations on document metadata recovery .........................
Better performance for large SharePoint farms ................................................
Performing SharePoint 2007 granular backups................................................
Configuring a SharePoint 2007 Client resource for granular backup ...........
SharePoint 2007 granular recovery ....................................................................
SharePoint recovery search .................................................................................
Searching and marking tips ..........................................................................
Searching for a SharePoint item ...................................................................
Performing a granular recovery of SharePoint items ......................................
SharePoint prerecovery options..........................................................................
Selecting alternate destination options........................................................
SharePoint recovery options ...............................................................................
Commands for SharePoint 2007 granular backup and recovery...................
nsr_moss_save.exe command.......................................................................
nsr_moss_recover.exe command .................................................................
Using administrator account other than that created during
installation ..............................................................................................................
Chapter 8
170
170
170
171
171
172
173
175
177
177
178
179
179
181
182
182
186
188
Microsoft Exchange Server Backup and Recovery
Exchange Server backup and recovery..............................................................
Microsoft Exchange Server MAPI client and Collaboration
Data Objects support......................................................................................
Exchange Server versions supported by NMM client ..............................
NMM and Exchange Server support for L18N ..........................................
Exchange Server backup and recovery types.............................................
Updating and modifying Exchange credentials ........................................
Compatibility with previous Exchange Server backups ..........................
Exchange Server data backed up by NMM ................................................
Components used by NMM for Exchange Server backup
and recovery....................................................................................................
Specifying Exchange save sets for application data ..................................
Exchange application information variable settings .................................
Serverless backup ...........................................................................................
Snapshot process in an Exchange Server production and
proxy client setup ...........................................................................................
Best practices for Exchange backup and recovery.....................................
Performing Exchange Server backups...............................................................
Configuring Exchange Server backups .......................................................
Configuring an Exchange Client resource ........................................................
Configuring an Exchange Client resource ..................................................
Configuring backup for Microsoft Exchange Server 2007 CCR ..............
Performing Exchange Server recovery ..............................................................
Exchange recovery requirements and limitations .....................................
Quick recovery of a deleted mailbox ...........................................................
Exchange recovery options ...........................................................................
Roll-forward recovery....................................................................................
Point-in-time recovery ...................................................................................
Mounting the database after recovery.........................................................
Rollback restore of Exchange data by using NMM ...................................
Exchange rollback operation requirements ................................................
Recovery to storage groups ................................................................................
190
190
190
190
190
191
192
193
193
194
196
197
198
199
202
202
204
204
206
215
215
216
217
218
219
220
220
221
222
7
Contents
Selecting a storage group restore destination ............................................. 222
Recovery to the original storage group ....................................................... 223
Recovery to an Exchange RSG ...................................................................... 224
Selecting and recovering items in the RSG database................................. 229
Recovery to an alternate storage group ....................................................... 231
Microsoft Exchange Server 2007 directed recovery.......................................... 234
Performing a directed recovery .................................................................... 234
Microsoft Exchange Server 2007 disaster recovery .......................................... 237
Microsoft Exchange Server 2007 (standalone) disaster recovery............. 237
Microsoft Exchange Server 2007 disaster recovery for Exchange
CCR to a 2 node cluster in a production environment.............................. 238
Microsoft Exchange Server 2007 disaster recovery for Exchange
CCR in a production environment ............................................................... 240
Chapter 9
Microsoft Data Protection Manager Backup
Data Protection Manager backup and recovery ............................................... 244
DPM Server versions supported by NMM client....................................... 245
DPM backup and recovery types ................................................................. 246
DPM data backed up by NMM..................................................................... 246
Transportable backups of DPM replicas not supported ........................... 246
Components used by NMM for DPM backup and recovery ................... 247
Specifying DPM save sets for application data .......................................... 248
DPM application information variable settings ......................................... 250
Performing DPM Server backups ....................................................................... 251
Configuring DPM backups............................................................................ 251
Configuring a DPM Client resource ................................................................... 252
Performing DPM Server recovery ...................................................................... 254
Chapter 10
Microsoft Data Protection Manager Granular and Disaster Recovery
Performing a granular recovery of DPM objects .............................................. 256
DPM Recovery Options Summary ............................................................... 259
Performing a granular directed recovery to a DPM-protected server........... 260
DPM directed granular recovery scenario .................................................. 260
Performing a granular directed recovery .................................................... 261
DPM disaster recovery ......................................................................................... 263
Chapter 11
Microsoft Hyper-V Backup and Recovery
Hyper-V backup and recovery ............................................................................ 268
Hyper-V version requirements ..................................................................... 268
Operating system versions supported by NMM client ............................. 268
Hyper-V parent partition backup and recovery types .............................. 269
Hyper-V parent partitions, child partitions, and applications
within child partitions.................................................................................... 269
Hyper-V storage and backup options supported by NMM ..................... 270
Backup roadmap for Hyper-V ...................................................................... 273
Best practices for Hyper-V parent partition backup and recovery.......... 275
Best practices for Hyper-V child partition applications backup
and recovery ................................................................................................... 276
Components used by NMM for Hyper-V backup and recovery ............. 277
Specifying Hyper-V save sets for application data .................................... 277
Hyper-V application information variable settings ................................... 279
Performing Hyper-V backups ............................................................................. 280
Preparing a virtual machine for backup...................................................... 280
8
Contents
Hyper-V configuration requirements for backing up a virtual
machine that contains multiple volumes ....................................................
NMM support for proxy backup of Hyper-V ............................................
Configuring Hyper-V backups .....................................................................
Configuring a Hyper-V Client resource ............................................................
Performing Hyper-V recovery ............................................................................
Off-line and on-line recovery........................................................................
Restrictions for backup and recover of Hyper-V virtual
machines in a Windows 2008 Failover Cluster ..........................................
Hyper-V recovery to the original machine and location ..........................
Hyper-V recovery to a different machine or location ...............................
Hyper-V parent partition disaster recovery .....................................................
Chapter 12
284
285
285
288
Microsoft Windows Cluster Backup and Recovery
Windows Server Cluster backup and recovery................................................
Cluster support in NMM client ....................................................................
Compatibility with previous Windows Cluster backups.........................
Components used by NMM for Windows Cluster backup
and recovery....................................................................................................
Windows Cluster application information variable settings ...................
Performing Windows Server Cluster backups .................................................
Cluster failover and backups ........................................................................
Windows Server Cluster recovery......................................................................
Windows Server 2003 clustered client disaster recovery ...............................
Disaster recovery of a cluster client .............................................................
Disaster recovery from a legacy VSS backup .............................................
Disaster recovery of a clustered client from a legacy non-VSS
backup ..............................................................................................................
Monitoring a recovery ...................................................................................
Chapter 13
280
280
281
282
284
284
292
292
292
292
293
294
294
295
295
295
298
300
301
Microsoft Active Directory Backup and Recovery
Active Directory backup and recovery..............................................................
Types of supported backup and recovery ..................................................
Performing an Active Directory or ADAM backup ........................................
Planning backups for disaster recovery ......................................................
Performing a granular Active Directory or ADAM backup...........................
Considerations for granular backup............................................................
Configuring Active Directory or ADAM granular backups ....................
Task 1: Set up local Windows SYSTEM account as an ADAM
administrator...................................................................................................
Task 2: Configure a pool for backup operations ........................................
Task 3: Configure a backup schedule ..........................................................
Task 4: Configure a backup group...............................................................
Task 5: Configure a Client resource .............................................................
Task 6: Configure privileges .........................................................................
Recovering all Active Directory data.................................................................
Performing a granular recovery of Active Directory objects and
attributes .................................................................................................................
Recovery Restrictions for Active Directory ................................................
Active Directory disaster recovery.....................................................................
304
304
305
305
306
306
307
307
308
310
310
311
313
314
316
317
320
9
Contents
Chapter 14
Snapshot Management
Roll over a snapshot.............................................................................................. 322
Roll back a snapshot ............................................................................................. 323
Rollback recovery requirements ................................................................... 323
Storage array descriptor file .......................................................................... 323
System rollback may fail with CLARiiON if the configuration
information is not current.............................................................................. 325
Performing a rollback recovery..................................................................... 325
Delete a snapshot .................................................................................................. 326
Chapter 15
NetWorker Client Management
Specifying ranges for service and connection ports......................................... 328
Specifying a password.......................................................................................... 329
Chapter 16
Troubleshooting
The troubleshooting process................................................................................ 332
Setting up notifications .................................................................................. 333
Finding errors .................................................................................................. 333
NMM client error messages................................................................................. 334
E_VOLUME_NOT_SUPPORTED ................................................................ 334
E_VETO_PROVIDER ..................................................................................... 334
VSS_E_WRITERERROR_RETRYABLE error code 0x800423f3 ................ 334
VSS CLIENT... Invalid Writer Selection... for APPLICATIONS .............. 335
access to the path is denied............................................................................ 335
Object reference not set to an instance of an object.................................... 335
NMM client issues................................................................................................. 336
Savegroup failed in scheduled backup........................................................ 336
In progress backup fails ................................................................................. 336
NetWorker is blocked by Windows Firewall.............................................. 336
Need to run utility to obtain Exchange Server information ..................... 336
Requesting a snapshot for a volume on hardware that is
not transportable ............................................................................................. 337
PowerSnap service must be restarted after the NetWorker service
is restarted ........................................................................................................ 337
PowerSnap validation for recovery fails ..................................................... 338
Snapshot backup may fail with writer errors ............................................. 338
Microsoft I/O write error when taking a replica in a CLARiiON
array .................................................................................................................. 338
Creating a replica takes a long time in CLARiiON environment ............ 338
Cannot recover Exchange database due to overwrite prevention
error................................................................................................................... 338
Incremental backups within NetWorker Module for Exchange
Client may be promoted to full backups ..................................................... 339
Incremental backups within NetWorker Module for SQL Server
Client may be promoted to full backups ..................................................... 339
Diagnosing VSS Writer Issues....................................................................... 340
Modifying the Exchange 2007 COM+ component’s user account
or password ..................................................................................................... 340
Save sets may not appear in the recover window...................................... 341
Recovery of AES-encrypted files fails without passphrase ...................... 341
Unable to restore files encrypted with both Microsoft Windows
Encrypting File System and AES encryption.............................................. 341
Anti-virus programs block recovery ............................................................ 342
10
Contents
Checking log files..................................................................................................
NMM client......................................................................................................
PowerSnap client ............................................................................................
Replication Manager ......................................................................................
Solutions Enabler............................................................................................
Active Directory..............................................................................................
NetWorker server ...........................................................................................
Manually stopping and starting services ..........................................................
Other troubleshooting resources ........................................................................
343
343
343
344
344
345
345
346
347
Glossary
Index
11
Contents
12
Figures
Title
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Page
Simple LAN-based configuration ....................................................................................... 26
LAN-based configuration with a proxy client .................................................................. 27
LAN-free configuration........................................................................................................ 28
VSS backup process .............................................................................................................. 30
NMM client main window .................................................................................................. 39
Exclusion representation of recover items ........................................................................ 40
Snapshot Management view ............................................................................................... 41
Monitor view ......................................................................................................................... 42
Selected and partially selected items.................................................................................. 44
Local client on the taskbar next to Client......................................................................... 106
Configuration Options dialog box.................................................................................... 106
Select Viewable Clients dialog box................................................................................... 107
NMM client dropdown list ................................................................................................ 107
SharePoint 2007 stand-alone farm configuration ........................................................... 134
SharePoint 2007 distributed farm configuration ............................................................ 135
Restore Remote Dependencies dialog box ...................................................................... 150
Newly created databases created in SharePoint............................................................. 164
NMM backups available for recovery.............................................................................. 165
SQL databases selected for recovery in NMM................................................................ 165
Databases recovered to SQL Server.................................................................................. 166
SharePoint Server Central Administration window...................................................... 166
SharePoint Portal Server Central Administration, Manage Portal Sites page ........... 167
Portal creation and restore options................................................................................... 167
Portal restore and database names ................................................................................... 168
Confirmation page for request to create a portal site .................................................... 168
Status page for portal creation process ............................................................................ 169
Operation Successful page displayed when portal creation is complete.................... 169
Recovered databases and SharePoint portal visible and available in SharePoint ..... 169
NMM displaying SharePoint 2007 browsing for recovery items................................. 179
NMM displaying Search tab for SharePoint 2007 recovery .......................................... 180
Sample output from nsr_moss_save ................................................................................ 187
Snapshot process with Exchange production server and proxy client ....................... 203
Select Storage Group Restore Destination dialog box ................................................... 226
Available and unavailable storage groups...................................................................... 230
Restore To command on context menu ........................................................................... 230
RSG Configuration error message .................................................................................... 231
RSG Configuration correct message................................................................................. 231
Public folders cannot be recovered to an RSG ................................................................ 232
Selected and partially selected items................................................................................ 234
Recovered items folders in user mailbox......................................................................... 235
13
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14
Local client on the taskbar next to Client .........................................................................
Configuration Options dialog box ....................................................................................
Select Viewable Clients dialog box ...................................................................................
NMM client dropdown list ................................................................................................
NMM protection of DPM ...................................................................................................
Directed granular recovery to a DPM-protected client..................................................
DPM recovery modes..........................................................................................................
Recover submenu for DPM Recovery ..............................................................................
DPM granular recovery available to file level.................................................................
DPM granular recovery options........................................................................................
Directed recovery of FileServer1 .......................................................................................
DPM database and replicas available as selectable items..............................................
DPM recovery modes..........................................................................................................
Recover submenu for DPM recovery ...............................................................................
DPM disaster recovery options .........................................................................................
Windows Server 2008 host with Hyper-V virtual machine child partitions ..............
Hyper-V storage options ....................................................................................................
NMM backup of Hyper-V components ...........................................................................
Disaster recovery order for a Hyper-V server.................................................................
Troubleshooting process.....................................................................................................
239
239
239
240
249
260
261
261
262
262
264
267
267
268
268
274
275
281
292
336
Tables
Title
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Page
Currently supported writers ............................................................................................... 31
Support for disaster and granular backup and recovery ................................................ 38
Backup tasks .......................................................................................................................... 48
Access privileges needed for NMM ................................................................................... 49
Best practices and considerations for application backups ............................................ 50
Conventional backup settings ............................................................................................. 57
Instant snapshot policy ........................................................................................................ 57
Instant snapshot with rollover policy ................................................................................ 57
Considerations for NMM Client backup schedules......................................................... 58
Save set syntax....................................................................................................................... 62
Special characters and their URL-encoded values ........................................................... 64
Application information variable settings......................................................................... 65
Configuration requirements for a proxy client................................................................. 67
Additional steps and considerations for backing up a clustered Client resource ....... 69
NMM System data commands............................................................................................ 83
Access privileges needed for NMM ................................................................................... 86
SQL save set syntax............................................................................................................. 115
Special characters and their URL-encoded values ......................................................... 117
SQL variable settings for Application Information attribute ....................................... 118
SQL Server backup and recovery best practices and considerations .......................... 122
Comparison of VSS, optimized, and granular backups ................................................ 131
SharePoint save set syntax ................................................................................................. 137
SharePoint application information variable settings ................................................... 139
Save sets for backup............................................................................................................ 140
SQL save set syntax for SharePoint 2003 ......................................................................... 153
SQL variable settings for application information attribute......................................... 156
SQL Server backup and recovery best practices and considerations .......................... 161
SharePoint Server backup list for disaster recovery ...................................................... 170
SharePoint recovery search fields ..................................................................................... 181
SharePoint granular recovery save set syntax for SharePoint 2007 objects................ 186
Application information variable settings....................................................................... 189
nsr_moss_recover command options............................................................................... 190
Exchange save set syntax ................................................................................................... 198
Exchange application information variable settings...................................................... 200
NMM Data Protection Manager commands................................................................... 251
DPM Server 2007 save set syntax...................................................................................... 252
DPM application information variable settings ............................................................. 254
15
41
42
43
44
45
46
47
48
49
50
51
52
16
Hyper-V child partition configurations ...........................................................................
Backup tasks for Hyper-V ..................................................................................................
NMM Hyper-V commands ................................................................................................
Hyper-V save set syntax .....................................................................................................
Hyper-V application information variable settings........................................................
Windows Cluster application information variable settings ........................................
Additional steps and considerations for backing up a clustered Client resource .....
Considerations for NMM client backup schedules ........................................................
System-only attributes that are not backed up................................................................
Retained attributes after object is deleted ........................................................................
Notifications and errors......................................................................................................
Services used in NMM client .............................................................................................
275
277
281
282
283
297
298
314
321
322
337
350
Preface
As part of an effort to improve and enhance the performance and capabilities of its product
lines, EMC periodically releases revisions of its hardware and software. Therefore, some
functions described in this document may not be supported by all versions of the software or
hardware currently in use. For the most up-to-date information on product features, refer to
your product release notes.
If a product does not function properly or does not function as described in this document,
please contact your EMC representative.
Audience
This guide is part of the EMC NetWorker Module for Microsoft Applications
documentation set, and is intended for use by system administrators during the setup
and maintenance of the product.
Readers should be familiar with the following technologies used in backup and
recovery:
Related
documentation
Conventions
◆
Storage subsystems, such as EMC CLARiiON or Symmetrix, if used
◆
EMC NetWorker software
◆
EMC NetWorker Snapshot management
◆
Microsoft Volume Shadow Copy Service (VSS) technology
Related documents include:
◆
EMC NetWorker Module for Microsoft Applications Release 2.2 Installation Guide
◆
EMC NetWorker Module for Microsoft Applications Release 2.2 Release Notes
◆
EMC NetWorker Release 7.5 Service Pack 1 Multiplatform Version Administration
Guide
◆
EMC Information Protection Software Compatibility Guide
◆
EMC Solutions Enabler Symmetrix CLI, Version 6.3 Quick Reference
EMC uses the following conventions:
Note: A note presents information that is important, but not hazard-related.
17
Preface
CAUTION
A caution note contains information that is essential to avoid data loss.
Typographical conventions
EMC uses the following type style conventions in this document:
Normal
Used in running (nonprocedural) text for:
• Names of interface elements (such as names of windows, dialog boxes, buttons,
fields, and menus)
• Names of resources, attributes, pools, Boolean expressions, buttons, DQL
statements, keywords, clauses, environment variables, filenames, functions,
utilities
• URLs, pathnames, filenames, directory names, computer names, links, groups,
service keys, file systems, notifications
Bold:
Used in running (nonprocedural) text for:
• Names of commands, daemons, options, programs, processes, services,
applications, utilities, kernels, notifications, system call, man pages
Used in procedures for:
• Names of interface elements (such as names of windows, dialog boxes, buttons,
fields, and menus)
• What user specifically selects, clicks, presses, or types
Where to get help
Italic:
Used in all text (including procedures) for:
• Full titles of publications referenced in text
• Emphasis (for example a new term)
• Variables
Courier:
Used for:
• System output, such as an error message or script
• URLs, complete paths, filenames, prompts, and syntax when shown outside of
running text.
Courier bold:
Used for:
• Specific user input (such as commands)
Courier italic:
Used in procedures for:
• Variables on command line
• User input variables
<>
Angle brackets enclose parameter or variable values supplied by the user
[]
Square brackets enclose optional values
|
Vertical bar indicates alternate selections - the bar means “or”
{}
Braces indicate content that you must specify (that is, x or y or z)
...
Ellipses indicate nonessential information omitted from the example
EMC support, product, and licensing information can be obtained as follows.
Product information — For documentation, release notes, software updates, or for
information about EMC products, licensing, and service, go to the EMC Powerlink
website (registration required) at:
http://Powerlink.EMC.com
Technical support — For technical support, go to EMC Customer Service on
Powerlink. To open a service request through Powerlink, you must have a valid
support agreement. Please contact your EMC sales representative for details about
obtaining a valid support agreement or to answer any questions about your account.
18
Preface
Your comments
Your suggestions will help us continue to improve the accuracy, organization, and
overall quality of the user publications. Please send your opinion of this document to:
[email protected]
19
Preface
20
1
Overview
This chapter includes the following topics:
◆
◆
◆
◆
◆
◆
◆
NetWorker Module for Microsoft Applications .......................................................
Supported configurations ............................................................................................
The VSS snapshot creation process.............................................................................
NMM client snapshot backups ...................................................................................
NMM client recoveries .................................................................................................
The NMM client interface ............................................................................................
NMM client graphical user interface..........................................................................
Overview
22
24
29
34
36
39
43
21
Overview
NetWorker Module for Microsoft Applications
The EMC® NetWorker® Module for Microsoft Applications (NMM) works with
Microsoft Volume Shadow Copy Service (VSS) technology to provide snapshot
backup and recovery services for file systems, and application, and operating system
data.
The NMM client allows for the creation of point-in-time snapshot data. A snapshot
can be retained on storage volumes for quick access. One can also perform a rollover
of a snapshot to a traditional backup medium such as a tape, file type device, or
advanced file type device. Data can be recovered either from a snapshot or from the
backup.
NetWorker Module features
This NetWorker Module can be set up in a LAN-based or LAN-free environment.
NMM provides the following features:
◆
Support for the following topologies:
• “Simple LAN-based configuration” on page 25
• “LAN-based configuration with a proxy client” on page 26
• “LAN-free configuration” on page 27
◆
Support for several types of backup and recovery, depending on the types
available for the given application or operating system:
• “Instant backup with or without rollover” on page 34
• “Nonpersistent rollover backup” on page 34
• “Serverless backup” on page 34
• “Conventional recovery” on page 36
• “Instant recovery” on page 36
• “Rollback recovery” on page 36
◆
Deduplication support for most of the Microsoft applications and Windows
operating systems that are supported by NMM:
• “Windows OS, application, and configuration support” on page 74 provides
detailed lists of applications and versions supported.
• For the NetWorker server, NetWorker Release 7.5 SP1 or later is required.
• For the Avamar server hardware, Avamar Axiom v4.0.2 or later is required.
◆
Application support:
• Microsoft Exchange Server 2007
• Microsoft Exchange Server 2003
• Microsoft SQL Server 2008
• Microsoft SQL Server 2005
• Microsoft Office SharePoint Server 2007
• Microsoft SharePoint Portal Server 2003
• Windows SharePoint Services 3.0
• Windows SharePoint Services 2.0
• Microsoft System Center Data Protection Manager 2007
22
Overview
• Windows Server 2008 with Hyper-V
• Windows Server 2008, Server Core Installation, with Hyper-V
• Windows volumes and system components
• Windows Cluster support
• Active Directory Application Mode (ADAM)
• Active Directory
Note: Application support often requires a minimum specific service pack level. The service
pack levels frequently change. The EMC Information Protection Software Compatibility Guide
provides the most up-to-date information about which service packs are required for each
application.
◆
Support for Windows VSS Provider:
• Microsoft Software Shadow Copy provider for use with most leading
hardware arrays
• EMC Symmetrix® and Symmetrix DMXTM arrays with EMC VSS Provider
• EMC CLARiiON® arrays with EMC VSS Provider
• Celerra® arrays with Celerra VSS Provider
• EqualLogic PS Series arrays with EqualLogic Integration ToolKit
• IBM Storage System DS8000
• IBM Storage System DS6000
The EMC Information Protection Software Compatibility Guide provides the most
up-to-date information about hardware array versions supported and NMM
compatibility.
Privileged user-level access
All NMM users must have administrator-level privileges for all resources they access:
◆
NetWorker server
◆
Each NetWorker Module for Microsoft Applications (NMM) client machine
◆
All Microsoft applications protected by NMM, such as:
• Windows Server
• Data Protection Manager (DPM)
• SharePoint Server
• SQL Server
• Exchange Server
◆
Network domains
◆
Storage servers
Windows and Microsoft application updates
Keep NMM client machines current with Microsoft Windows updates for the
operating system, as well as updates and hotfixes for each installed Microsoft
application protected by NMM. Otherwise NMM may not function correctly, or
backup or recovery may fail. For example, the NMM UI may fail if the latest .NET
updates are not installed.
NetWorker Module for Microsoft Applications
23
Overview
Supported configurations
This section describes the basic configurations in which the NMM client is supported.
Supported NetWorker Server hosts
The NMM client supports NetWorker Server 7.3.3 or later and NetWorker Storage
Node 7.3.3 or later on the following hosts:
◆
AIX
◆
HP-UX (PA-RISC)
◆
Solaris (SPARC)
◆
Windows
◆
Linux (NetWorker Server 7.4 SP3 or later and NetWorker Storage Node 7.4 SP3 or
later)
The NMM client supports Internationalization (I18N).
NetWorker 7.4 SP3 or later is required for Microsoft Office SharePoint 2007 granular
protection.
This release of NMM supports NetWorker cloning and staging to create and manage
additional copies of save sets. This NMM support requires NetWorker 7.5 or later.
The EMC NetWorker 7.5 Administration Guide provides details for cloning and staging.
The following NetWorker features are not supported:
◆
Localization (L10N)
◆
Installation of dedicated storage node on NMM client host.
Note: NMM does support proxy storage node, as described in “LAN-free configuration”
on page 27.
◆
Manual (adhoc) backups
Windows operating system support
NMM client supports the following versions of Windows and Windows-related
features:
24
◆
Microsoft Windows Server 2008 (x86, x64)—Standard, Enterprise, DataCenter
Editions
◆
Microsoft Windows Server 2003 R2 (x86, x64)—Standard, Enterprise, DataCenter
Editions
◆
Microsoft Windows Server 2003 (x86, x64)—Standard, Enterprise, DataCenter
Editions
◆
Microsoft Windows Storage Server 2003
◆
Microsoft Windows Storage Server 2003 R2
Overview
◆
System state recovery, file servers, and operating system roles including Active
Directory, DHCP, and Terminal Services
◆
GUID Partition Table disks:
• No disk size limit on Windows environments with CLARiiON storage
(SnapView™ clone, SnapView snap, and SAN Copy™ technologies only)
• Disk size limit of less than 2 TB on Windows environments with Symmetrix
storage
Note: The following Windows Server versions are not supported: Windows IA64 editions.
In addition, the following Windows features are not supported:
◆
BitLocker encryption
◆
LAN-based Proxy Client or LAN-free backups, if dynamic disks are used
◆
VSS Hardware Providers with Windows dynamic disks
◆
Microsoft Software Shadow Copy provider to perform persistent snapshots of
clustered disks
◆
Windows Automated System Recovery (ASR)
Note: Windows Storage Server Single Instance Storage (SIS) is supported, but recoveries
will recover file data for all duplicate files. Data will not be lost if there is sufficient disk
space to hold the duplicate copies.
The EMC Information Protection Software Compatibility Guide contains additional and
the most up-to-date information about NMM compatibility.
Simple LAN-based configuration
Figure 1 on page 26 shows a simple LAN-based configuration with a storage area
network. In a LAN-based configuration, the production data is stored on a storage
area network (SAN) or direct-attached storage (DAS). The data path to the
NetWorker storage node devices is over a LAN. The application server that is being
protected has the NMM client installed.
The data moves as follows:
1. The NetWorker Server, which must be release 7.3.3 or later, initiates the process
by contacting the application server where the NMM client software is installed.
2. The application server with the NMM client software creates a snapshot of the
data on the storage volume.
Note: In Figure 1 on page 26, snapshots are represented by S1 through S5.
3. The application server with the NMM client software uses a snapshot in primary
storage to transfer the data over the LAN and into a conventional backup
medium such as a file type disk, advanced file type disk, or tape.
25
Overview
NetWorker
Server 7.3.3 or later
LAN
Application server
NetWorker Module for
Microsoft Applications
NetWorker
Storage
Node 7.3.x
SAN
Primary storage
(For example,
CLARiiON or Symmetrix)
S5
S1
S4
Traditional
storage
S3
S2
GEN-000747
Figure 1
Simple LAN-based configuration
LAN-based configuration with a proxy client
Figure 2 on page 27 shows a LAN-based configuration with a storage area network
and a proxy client. Both the application server and the proxy client have the NMM
client installed.
Note: The application server and the proxy client must use the same version, NetWorker
Module for Microsoft Applications release 2.2. If NMM 2.2 is installed on either the application
server or the proxy client machine, then the other machine must also use NMM 2.2. Backup
and recovery will fail if NMM 2.2 is used on one of these machines, and an earlier version of
NMM is used on the other.
A proxy client is a host that acts as a remote data mover (DM) when snapshots are
rolled over to a conventional backup medium. A proxy client frees resources on the
application server. When a backup operation uses a proxy client, it is known as a
severless backup.
1. The NetWorker server initiates the process by contacting the application server
where the NMM client software is installed.
26
Overview
3. The snapshot is made visible to the proxy client.
4. The proxy client uses a snapshot in primary storage to transfer the data over the
LAN and into a conventional backup medium such as a file type disk, advanced
file type disk, or tape.
NetWorker Server
7.3.3 or later
LAN
Application server
Proxy client
SAN
Primary storage
(For example,
S4
S5
S1
S3
NetWorker
Storage
Node 7.3.x
Traditional
storage
S2
GEN-000748
Figure 2
LAN-based configuration with a proxy client
LAN-free configuration
Figure 3 on page 28 shows a LAN-free configuration with a storage area network.
The NMM client is installed on the application server. The proxy client is installed on
the NetWorker storage node.
You can set up a NetWorker storage node release 7.3.3 or later as a proxy client to
avoid the network traffic that is generated when a snapshot is rolled over to a
conventional backup medium. Network traffic is avoided because both the
conventional backup device and the snapshot are directly attached to the storage
node.
Note: Install the proxy client software by using the NMM client software package.
1. The NetWorker server initiates the process by contacting the application server
27
Overview
4. The proxy client, in this case the storage node, uses a snapshot in primary storage
to transfer the data into a conventional backup medium such as a file type disk,
advanced file type disk, or tape.
The NetWorker server and the application server communicate through the LAN.
However, the data itself is not transferred across the LAN because the
conventional backup medium is attached directly to the NetWorker storage node.
NetWorker Server
7.3.3 or later
LAN
Application server
Proxy client
NetWorker Storage
Node 7.3.x
SAN
Primary storage
(For example,
Traditional
storage
S4
S5
S1
S3
S2
GEN-000750
Figure 3
28
LAN-free configuration
Overview
The VSS snapshot creation process
Microsoft Volume Shadow Copy Service (VSS) acts as a coordinator among the
following three components that create, modify, back up, and recover data:
◆
Requestors
Requestors include the NetWorker backup application.
◆
Writers
Including writers for operating system data, application data such as Microsoft
Exchange Server 2003 or Microsoft SQL Server data, and file system volumes such
as a D:\ data drive.
◆
Providers
Providers include the Microsoft VSS software provider or hardware-based
providers such as the EMC VSS Provider.
In VSS, NetWorker software is a requestor—an application that performs the backup
and recovery of applications or services. When a requestor performs a backup or
recovery, the following occurs:
1. The requestor asks for information from VSS.
2. VSS reviews the request for validity.
3. VSS sends writer data from multiple application and system writers to the
requestor.
4. The requestor tells VSS which writers and volumes to include in the snapshot.
Each application and service that supports VSS has its own writer, which
understands how the application or service works. After the writer signals that it has
prepared the data, the following occurs:
1. The VSS software does the following:
a. Directs the writer to freeze input/output (I/O) to the selected volumes.
b. Queues transactions for later processing.
c. Calls a provider to capture the requested data.
2. The provider does the following:
a. Captures the prepared data.
b. Creates a snapshot that exists side-by-side with the live volume.
3. After the snapshot is created:
a. The provider signals VSS, this signal tells the writer to resume activity.
b. I/O is released to the selected volumes and any queued writes that arrived
during the provider's work are processed.
Figure 4 on page 30 illustrates the VSS backup process.
29
Overview
Requestor
1. NetWorker software (the requestor)
asks VSS to enumerate writers and
gether their metadata.
4. Requestor
asks VSS
to create
snapshot.
2. Writers provide an XML description
of backup components and define
the recover method.
Writer
Writer
Writer
Writer
5. VSS tells the writers to
freeze activity.
Writer
7. VSS tells the writers to
resume activity.
3. VSS asks which providers can
support a snapshot for each of
the required volumes.
Microsoft VSS
Software Provider
Disk 1
Disk 2
6. VSS tells the providers to
create the snapshot of the
current state on disk.
EMC VSS Provider
(CLARiiON or Symmetrix)
Disk 3
Disk 3
GEN-000188
Figure 4
VSS backup process
Writers
Writers provide information about what data to back up, and specific methods for
handling components and applications during backup and recovery. They also
identify the type of application or service that is being backed up.
If a service or application is present on a system but is not active, information from its
writer will not be available. Consequently, a writer can appear or disappear from
backup to backup.
30
Overview
Table 1 on page 31 lists the writers that are supported for the NMM client.
Table 1
Currently supported writers
Writer type
Writers supported by NMM
Volumes
A file system path such as D:\data
Note: NMM only supports NTFS volumes. It does not support FAT16 and
FAT32 volumes. Also, NMM does not support backup of raw volumes.
Applications
Microsoft Active Directory Application Mode (ADAM)
Microsoft Exchange Server 2003
Microsoft System Center Data Protection Manager 2007
Microsoft Distributed File Replication (Windows Server 2003)
Microsoft Active Directory Lightweight Directory Services (AD LDS)
Microsoft SharePoint Portal Server 2003
Microsoft Office SharePoint Server 2007
Windows SharePoint Services 2.0
Microsoft SQL Server 2008
Microsoft SQL Server 2005
Windows Server 2008 (Full Installation) with Hyper-V
Windows Server 2008 (Server Core Installation) with Hyper-V
Windows system components
(Windows Server 2003 and Windows
Server 2008 unless otherwise noted)
BITS (Background Intelligent Transfer Service)
COM+ REGDB (Registration Database)
Certificate Authority
Cluster Database (Windows Server 2003)
Cluster Writer (Windows Server 2008)
DHCP (Dynamic Host Configuration Protocol) Jet
Event log
FRS (File Replication Server)
FSRM (File Storage Resource Manager)
IIS Configuration Writer (Windows Server 2008)
IIS Metabase
MSSearch Service Writer (Windows Server 2008)
NPS VSS Writer (Network Policy and Access Service, Windows Server
2008)
NTDS (Active Directory Schema)
Registry
RSM (Removable Storage Manager)
Remote Storage Manager
Shadow Copy Optimization Writer (Windows Server 2008)
SQL Embedded (SQL 2005 Express, or SQL 2008 Express)
Terminal Services Gateway Writer (Windows Server 2008)
Terminal Services Licensing
WINS (Windows Internet Name Service)
WMI (Windows Management Instrumentation)
Note: Files located in an FTP directory are not included in a SYSTEM COMPONENTS
snapshot. Typically, FTP files are stored by IIS in C:\inetpub. IIS Metabase files are
configuration files, which are backed up as part of SYSTEM COMPONENTS. The FTP files are
content files, and must be backed up separately.
31
Overview
Providers
A provider is used to capture snapshots. One of the following types of providers are
available:
◆
EMC VSS Hardware Provider, which supports EMC Symmetrix, Symmetrix
DMX, and CLARiiON arrays
◆
EMC Celerra VSS Hardware Provider, which supports Celerra arrays
◆
Dell EqualLogic VSS Hardware Provider, which supports EqualLogic PS Series
arrays
◆
Microsoft Software Shadow Copy provider
◆
IBM SystemStorage DS Open API Package, which supports IBM Storage, which
supports IBM Storage System DS6000 and DS8000
The Microsoft Software Shadow Copy provider is a host-based provider that works
with any type of storage hardware. The EMC VSS Hardware Provider works with an
CLARiiON or a Symmetrix disk array. Some features of the NMM client, such as
serverless backup and rollback recovery, are available only with hardware providers
such as the EMC VSS Hardware Provider.
The NetWorker Module for Microsoft Applications Release 2.2 Installation Guide provides
more information about using NMM with a VSS Hardware Provider, including
hardware array installation and configuration, and NMM installation and
configuration.
Note: NMM does not support rollback restore for backups performed by EqualLogic or Celerra
hardware. Only conventional restore and snapshot restore are supported for backups
performed by EqualLogic or Celerra hardware.
Dynamic volume support
In Windows Server 2003 and Windows Server 2008:
◆
NMM supports snapshots of dynamic volumes that use the Microsoft Software
Shadow Copy provider.
◆
NMM provides software-based support for:
• Dynamic volume on a single disk.
• Dynamic volume spanning multiple disk.
◆
NMM does not support dynamic volumes in striping.
◆
NMM does not support hardware-based VSS transportable snapshots of dynamic
volumes. NMM will attempt to perform a hardware-based VSS transportable
snapshot when a proxy backup host is specified by using NSR_DATA_MOVER in
the application information field of the Client resource.
For example, hardware-based transportable snapshots are not supported when
the dynamic volumes are on CLARiiON storage. NMM will attempt to perform
the hardware-based snapshot, but the operation will fail during import of the
hardware-based snapshot on the proxy backup host.
!
IMPORTANT
Any attempt to take a hardware-based replica of a dynamic volume will fail, and
cause instability in the system.
32
Overview
Windows Server 2008 introduces support for local, hardware-based snapshots of
dynamic volumes. NMM does not support local, nontransportable, hardware-based
snapshots of dynamic volumes on Windows Server 2008. For customer environments
where dynamic volumes are provisioned by using a storage array, customers must
not install the storage array vendor’s VSS provider on the production host. For
example, if a dynamic volume was created on a CLARiiON LUN, customers must not
install the EMC VSS Hardware Provider on the production host.
Providers and backup performance
The Microsoft VSS software provider uses a copy-on-write snapshot technology.
Some hardware providers, such as the EMC VSS Hardware Provider, support
split-mirror snapshot technology and copy-on-write technology. Although
copy-on-write technology requires less hardware resources, split-mirror technology
provides better performance.
Large or fast-changing databases should not be saved with copy-on-write snapshot
technology. The input/output requirements of these databases can cause the
copy-on-write cache to overflow and invalidate the snapshot. Additionally,
copy-on-write technology cannot match the performance of split-mirror technology.
For large or fast-changing databases, use the EMC VSS Hardware Provider in
conjunction with EMC hardware to provide a split-mirror storage solution such as a
CLARiiON clone or a Symmetrix BCV.
33
Overview
NMM client snapshot backups
Backups are configured as scheduled snapshot backups on a NetWorker server
release 7.3.3 or later.
Note: Manual (adhoc) backups, from either the command line or from the NMM client, are not
supported. However, one can manually start a scheduled backup at any time.
The NMM client supports three types of snapshot backups:
◆
“Nonpersistent rollover backup” on page 34
◆
“Instant backup with or without rollover” on page 34
◆
“Serverless backup” on page 34
Snapshot backups are specified through a snapshot policy and various settings for
the Client resource. Snapshot policies and Client resources are set up on the
NetWorker server. Chapter 2, “Configuring a Scheduled Backup,” provides more
information about configuring snapshot policies and Client resources.
Nonpersistent rollover backup
A nonpersistent snapshot backup (also referred to as a live backup) creates a
point-in-time snapshot of the data and then performs a rollover to a conventional
backup medium such as tape, file, or advanced file type device. The original snapshot
is automatically deleted from the NMM client.
Instant backup with or without rollover
An instant backup creates and retains a snapshot on a mounted storage disk volume.
Depending on how backups are configured, a snapshot that is created during an
instant backup may be configured to roll over to a conventional backup. While the
snapshot resides on a mounted storage disk volume, it is referred to as a persistent
snapshot.
Retention of persistent snapshots on a mounted storage disk volume enables one to
perform an instant recovery. The number of persistent snapshots that can be retained
on a volume depends on the amount of resources available. Other factors that
constrain the number of persistent snapshots include:
◆
Snapshot technology that is used, either copy-on-write or split-mirror.
◆
Input/output patterns of the drive.
◆
Processing involved in running a snapshot backup group.
Serverless backup
A proxy client is used in conjunction with instant snapshot backups to provide
serverless backups. Serverless backups free the application server from much of the
processing involved in a nonpersistent or instant snapshot backup. In a serverless
backup, a snapshot is created on the application server and then immediately moved
to the proxy client for further processing.
Serverless backups are particularly useful when there is additional processing
involved in a snapshot backup. For example, to help determine whether a snapshot of
a Microsoft Exchange database is consistent, the Exchange utility, eseutil, should be
34
Overview
run against the snapshot. Running eseutil can be disk intensive. Offloading the work
from the Exchange Server to a proxy client frees resources on the Exchange Server.
A serverless backup can be rolled over to a conventional backup. That backup can
then remain on the mounted storage disk volume as a persistent snapshot.
Note: Serverless backup is available only with VSS hardware providers and is subject to the
capabilities of the vendor’s hardware. The EMC VSS Provider for CLARiiON and Symmetrix
provides the capability to perform serverless backup operations for these storage platforms.
NMM client snapshot backups
35
Overview
NMM client recoveries
There are three types of recoveries:
◆
“Conventional recovery” on page 36
◆
“Instant recovery” on page 36
◆
“Rollback recovery” on page 36
By default, recoveries are performed from a conventional backup. If a conventional
backup is not available for the selected browse time, an instant recovery is performed.
The default recovery method can be specified in the NetWorker recovery options
dialog box. “NetWorker recovery options” on page 89 provides more information
about these recovery methods.
Conventional recovery
A conventional recovery consists of recovering data from a snapshot that has been
rolled over to tape. Conventional recoveries support the same level of item selection
as do instant recoveries.
Instant recovery
Instant recoveries are performed with persistent snapshots. An instant recovery can
take less time to complete than a conventional recovery because the snapshot is
available on a mounted disk storage volume rather than on a conventional backup
medium.
Instant recoveries support the selection of individual files in a file system snapshot.
Additionally, instant recoveries support the selection of individual components at
whatever level of granularity is supported by the system or application writer.
If the persistent snapshot is created by using a proxy client, the snapshot file system
on the proxy client is read to restore the files. In this way, performance is improved
even further because the job of processing the snapshot for recovery is off-loaded
from the NMM client production host to the proxy client.
The procedure to perform an instant recovery of file system data differs if only a
persistent snapshot is available. Chapter 4, “System Backup and Recovery,” provides
more information about performing recoveries.
Note: When performing an instant recovery, the snapshot may become invalidated. This can
occur if the snapshot is based on copy-on-write technology. If a large amount of data is
recovered to the original volume, the cache may overflow. This is true for the Microsoft VSS
software provider and for hardware providers that use copy-on-write technology, such as
CLARiiON snapshots and Symmetrix virtual devices.
Rollback recovery
A rollback recovery is the process of returning data volumes to an earlier point in
time. The entire contents of a persistent snapshot are recovered during a rollback
recovery. One cannot select individual components in a rollback recovery operation.
A rollback recovery is the fastest way to recover the contents of a persistent snapshot.
A rollback recovery is known as a "destructive" recovery because all changes that
were made to the file systems after the selected snapshot was taken are overwritten.
36
Overview
For example, suppose that a snapshot of a database is taken every hour. At 5:30 P.M.,
it is determined that the database is corrupt and that the last known consistent
database snapshot was taken at 5:00 P.M. When the database host is rolled back to the
5:00 P.M. snapshot, all database transactions between 5:00 P.M. and 5:30 P.M. are
overwritten.
The following list describes additional requirements and conditions that apply to
snapshots and rollback restore:
◆
The snapshot to be rolled back must reside on a CLARiiON or Symmetrix storage
system that is supported with the EMC VSS Provider.
◆
The NMM client to which the snapshot is to be rolled back must be supported
with a proxy client.
◆
NMM does not support rollback restore for backups performed with EqualLogic
or Celerra hardware. Only conventional restore and snapshot restore are
supported for those hardware products.
“Roll back a snapshot” on page 327 provides more information about rollback
recoveries.
Disaster and granular backup and recovery for applications
NMM provides disaster and granular backup and recovery for most applications:
◆
Full disaster backup and recovery means the entire volume or database for that
application is backed up, and the entire volume or database is recovered as a
whole.
◆
In disaster backup and recovery, items cannot be selected at a granular level.
◆
In granular backup, individual items can be selected for backup.
◆
In granular recovery, individual items can be selected for recovery.
For some applications, granular recovery of items requires that the items be specified
for a granular backup. For other applications, items cannot be specified for backup at
a granular level, but can be selected for granular recovery from the full backup.
NMM client recoveries
37
Overview
Table 2 on page 38 lists the support provided by NMM for disaster and granular
backup and recovery for Microsoft applications and volumes.
Table 2
38
Support for disaster and granular backup and recovery
Active
Directory
SQL
Server
Exchange
Server
Microsoft
Office
SharePoint
Server 2007
Disaster
(full
backup)
X
X
X
X
X
Granular
backup
X
X
X
X
Disaster
(full
recovery)
X
X
X
X
Granular
recovery
X
X
X
X
X
Data
Protection
Manager
Hyper-V
File
system
X
X
X
X
X
X
Overview
The NMM client interface
The NMM client interface has three views:
◆
“Recover view” on page 39
◆
“Snapshot Management view” on page 41
◆
“Monitor view” on page 42
Recover view
With the exception of rollback recovery, all recoveries are performed from the
Recover view. Rollback recoveries use the Snapshot Management view. Figure 5 on
page 39 shows the Recover view.
Main toolbar
Available views
Figure 5
Summary of selected items
Navigation tree
Client selection
(clustered client only)
View-specific toolbar
NMM client main window
39
Overview
The summary of selected items can also display exclusions, which shows only those
items that are not selected for recovery. This view can be helpful when many items
are selected for recovery. Figure 6 on page 40 shows an exclusion representation.
Items in red are
excluded from
recovery
Figure 6
Exclusion representation of recover items
To view an exclusion representation, right-click in the summary pane at the bottom of
the window and select Exclusion Representation.
40
Overview
Snapshot Management view
Various operations on snapshots are performed from the Snapshot Management
view shown in Figure 7 on page 41.
Figure 7
Snapshot Management view
The following operations can be performed on snapshots:
◆
Rollover snapshot
◆
Rollback snapshot
◆
Delete snapshot
In most cases, snapshot rollover and delete operations are performed automatically
based on snapshot policies, which are set up on the NetWorker server. However, if
you need to manually delete or roll over a snapshot, you can do so in the Snapshot
Management view. Chapter 2, “Configuring a Scheduled Backup,” provides more
information about snapshot policies. Chapter 14, “Snapshot Management,” provides
more information about performing manual snapshot operations from the Snapshot
Management view.
41
Overview
Monitor view
Recovery and snapshot management operations can be monitored in the Monitor
view shown in Figure 8 on page 42.
Figure 8
Monitor view
Most messages displayed in the Monitor view are also written to log files. Chapter 16,
“Troubleshooting,” provides more information about log files. You can also copy and
paste text from the Monitor view to another application.
42
Overview
NMM client graphical user interface
This section describes how to perform the basic tasks in the NMM client interface.
Opening the software and connecting to a NetWorker server
You must first connect to the NetWorker server on which the backup was configured
to recover data.
To connect to a NetWorker server:
1. From the Start menu, open the NMM client software:
• If you have opened the NMM client before, proceed to step 2 on page 43.
• If this is the first time you have opened the NMM client, the Change
NetWorker Server dialog box appears, and you can proceed to step 3 on
page 43.
2. From the Main toolbar, click the NetWorker Server icon.
The Change NetWorker Server dialog box appears.
3. Click Update Server List to browse for NetWorker 7.3.3 or later servers. The
discovery process may take a few minutes.
4. When the list is updated, select a server.
The selection appears in the Server field. You can also type a name in the Server
field.
5. Click OK.
Note: You can also select Configure Options from the Options menu to connect to a NetWorker
server.
43
Overview
Selecting an item for recovery
To select items for recovery, select the checkbox beside a node in the navigation tree.
A check mark indicates that the node is selected. By default, all items contained in the
node, such as folders and files, will also be selected for recovery.
To clear an item contained in a selected node, expand the node and clear the
checkbox beside the item. The check mark disappears.
Figure 9 on page 44 depicts one selected item and several partially selected nodes in
the navigation tree.
Check mark indicates that the item is selected.
Solid box indicates that
some items, not all, below
this node are selected
Figure 9
Selected and partially selected items
Searching for an item
To search for a recovery item:
1. Click Search above the navigation tree.
2. In the Path field, type a directory path.
3. Optionally, in the Name field, type the name of the search item. You can refine
your search as follows:
• Literal match (case-insensitive)
Type abc to return abc, ABC, or AbC but not abcd or ABCD.
• Literal match (case-sensitive)
Type "abc" to return abc but not ABc or abcd.
• Name contains (case-insensitive)
Type %abc% to return abc, abcd, ABCD, or xyzABCde.
• Name starts with (case-insensitive)
Type abc% to return abcd or ABCde but not xyzABCde.
44
Overview
• Name ends with (case-insensitive)
Type %abc to return xyzAbc but not ABCde.
• Single-character match search using the ? wildcard
Type ? to return single character entries and drive volumes, such as C or D.
Type WMI?Writer to return WMI Writer.
• Multiple-character match search by using the * wildcard
Type *.txt to return all entries with a .txt extension.
Type * to return all items within the selected container.
Type *writer* to return all writers.
• Search by using both the * and ? wildcard
Type *??I*writer* to return WMI Writer.
4. Click Search.
The Details pane displays the results of the search.
Specifying a recovery browse time
To browse for items that were backed up on or before a particular time:
1. From the application toolbar, click the Browse calendar icon.
2. Select the arrows to select the date and time, and then click OK.
The navigation tree displays backup items from the specified date and earlier.
Selecting a backup version for recovery
This topic applies to SharePoint granular recovery, DPM granular recovery, system
recovery, or Active Directory recovery operations only. Chapter 14, “Snapshot
Management,” provides more information about rollback snapshot recoveries.
To view the versions of an item that were backed up prior to the current browse time:
1. In the navigation tree, select a recover item.
“Selecting an item for recovery” on page 44 provides more information about
selecting an item.
2. Right-click the selected item and select Versions from the pop-up menu.
3. Select the backup version to recover.
4. Select Use selected item backup time as new browse time to change the browse
time to the same date as the selected backup version.
5. Click OK.
The recover items displayed in the navigation tree are from the selected browse
time.
45
Overview
Viewing the volumes required for a recovery
There are no NetWorker volumes associated with a snapshot backup until the
snapshot has been rolled over to a NetWorker server. If you select to view the
required volumes for a system component or application data snapshot that has not
been rolled over, no volumes will be displayed.
To view the volumes that are required for a recovery operation:
1. In the navigation tree, select a recover item.
selecting an item.
2. Right-click the selected item and select Required Volumes from the pop-up
menu.
The Required NetWorker Volumes dialog box appears with a list of the volumes
that must be mounted:
• If a volume is already mounted, the device name is shown after the volume
name.
• If a required volume is not mounted, the administrator is prompted to mount
it after the recovery operation is started.
3. Click OK.
46
2
Configuring a
Scheduled Backup
◆
◆
◆
◆
◆
◆
◆
◆
◆
◆
Backup configuration roadmap ..................................................................................
Task 1: Configure a backup pool for snapshot operations ......................................
Task 2: Configure snapshot policies ...........................................................................
Task 3: Configure a backup schedule .........................................................................
Task 4: Configure a backup group ..............................................................................
Task 5: Configure a Client resource ............................................................................
Task 6: Configure privileges ........................................................................................
Task 7: Configure a proxy client..................................................................................
Backing up a clustered NMM Client ..........................................................................
Moving an NMM Client to another NetWorker server ...........................................
48
52
55
58
59
61
66
67
69
70
47
Backup configuration roadmap
This chapter describes the seven tasks required to configure most types of backups.
Review Table 3 on page 48 to determine which tasks are required to configure your
backups.
Table 3
Backup tasks (page 1 of 2)
Type of backup
Required action
Windows volumes:
• Volume data such as a drive letter, for
example: “D:\”
• A file system path, for example “D:\data”
Windows system components, such as:
• The registration database
• Windows Cluster, etc.
• Other Window system components, as
listed in Table 1 on page 31.
• Complete tasks 1 through 7:
“Task 1: Configure a backup pool for snapshot operations” on
page 52
“Task 2: Configure snapshot policies” on page 55
“Task 3: Configure a backup schedule” on page 58
“Task 4: Configure a backup group” on page 59
“Task 5: Configure a Client resource” on page 61
“Task 6: Configure privileges” on page 66
“Task 7: Configure a proxy client” on page 67
Note: Specific Windows system components
do not have individual save sets. The
Windows system components are all backed
up under the save set “SYSTEM
COMPONENTS:\”
Application data, such as:
• Microsoft SQL Server
• Microsoft SharePoint
• Microsoft Exchange Server
• Microsoft System Center Data Protection
Manager (DPM)
• Microsoft Hyper-V
• Microsoft Windows Server Cluster
For example the save set
"APPLICATIONS:\SqlServerWriter"
“NetWorker Module features” on page 22
lists the specific versions supported for each
of these applications.
48
page 52
• Review the information in
“Best practices and considerations for backing up application
data” on page 50.
• Continue configuration with the specific instructions for the
application provided in:
“Performing SQL Server backups” on page 119
“Performing SharePoint 2003 backups” on page 157
“Performing Exchange Server backups” on page 206
“Performing DPM Server backups” on page 255
“Performing Hyper-V backups” on page 284
“Performing Windows Server Cluster backups” on page 298
Table 3
Backup tasks (page 2 of 2)
Type of backup
Required action
For all other applications and systems
supported by NMM
page 52
An NMM client that is part of a cluster
• Review the information in “Backing up a clustered NMM Client”
on page 69
page 52
An Active Directory server
Complete the tasks in Chapter 13, “Microsoft Active Directory
Backup and Recovery.”
Privileged user-level access
Because NMM requires access to protected data and system files, make sure all users
who will be running or configuring NMM backups and recoveries have the
appropriate level of privileges for all resources they must access. Table 4 on page 49
describes the privilege levels needed.
Table 4
Access privileges needed for NMM
Resource
Privileges needed
NetWorker Server
NetWorker Administrator or NetWorker Backup Operator
NetWorker Module for Microsoft
Applications client machine
Windows Administrator or Windows Backup Operator
All applications protected by NMM,
such as:
Windows Server
Microsoft Data Protection Manager
Microsoft Office SharePoint Server
Microsoft SQL Server
Microsoft Exchange Server
Windows Adminstrator
Network domain
Windows Domain Administrator
Storage node servers
Windows Administrator
49
Name resolution
NetWorker server and the NMM client machines need proper name resolution to
identify and back-translate computer names, such as from name-to-IP address or IP
address-to-name. Also, the NMM client uses the host machine NETBIOS or “short”
name when connecting to the NetWorker server to browse backups. If the NETBIOS
name is not found, NMM will not be able to display backups. To ensure clear
communication of computer names, make sure to perform the following:
◆
Add the NetWorker server name to either the local hosts file (located in
%SystemRoot%\system32\drivers\etc) or the Domain Name Server that contains
the names of all servers on your network.
◆
When configuring a Client resource, specify the NETBIOS name for the client in
the Aliases attribute. “Configuring a Client resource for System volumes,
components, and applications” on page 61 describes how to enter the NETBIOS
name in the Aliases attribute.
Best practices and considerations for backing up application data
Table 5 on page 50 lists the considerations for specifying values in the Save Set
attribute for a particular Client resource.
Table 5
50
Best practices and considerations for application backups (page 1 of 2)
Consideration
Best practice
Different policies for
application server data
For application servers such as SQL or Exchange Servers, back up the
server application data under a schedule different than the host operating
system data and volumes. Typically, application data is backed up several
times a day while operating system data and volumes are backed up less
frequently. To accomplish this, create a separate backup group, snapshot
policy, and Client resource for:
• Server application data
• Volumes and operating system data
Assign the appropriate snapshot policy and Client resource to each
backup group.
Installation path for
application server
program
Do not install application server program files on the same volume as the
application’s database files and log files.
Table 5
Best practices and considerations for application backups (page 2 of 2)
Consideration
Best practice
Copy-on-write versus
mirroring
Large or fast-changing databases should not be saved with copy-on-write
snapshot technology. Instead, use a split-mirror snapshot technology
such as a CLARiiON clone or a Symmetrix BCV.
Although copy-on-write technology requires less hardware resources,
split-mirror technology can better handle the requirements of large or
fast-changing databases.
Split-mirror technology is supported only with a VSS hardware provider,
such as the EMC VSS Provider. It is not provided by the built-in Microsoft
System Provider.
NetWorker modules and
the NMM Client
If you attempt to use both the NMM Client and a NetWorker module to
back up application data, the module backups will be promoted to Full
backups. “NMM client issues” on page 340 provides more information
about issues that may be encountered with the NMM client.
data
If SQL Server 2005 backups were performed when the SQL Server VSS
Writer service was disabled, and you subsequently enable the writer,
perform a complete backup of SQL 2005 data. Performing a complete
backup enables you to recover all SQL Server 2005 data from the same
application folder, which is named SQLServerWriter.
In addition to these best practices for all NetWorker Module for Microsoft
Applications (NMM) backups, there are application-specific best practices:
◆
“Best practices for Exchange backup and recovery” on page 203
◆
“Best practices for SQL Server backup and recovery” on page 121
◆
“Performing a recovery” on page 86
51
Task 1: Configure a backup pool for snapshot operations
Snapshots create metadata that is used in the management of snapshot operations
such as an instant recovery operation. Create a snapshot pool to store the save sets
that contain snapshot metadata.
Configuring a backup pool is a four-part process:
1. “Configuring a device” on page 52
2. “Configuring a label template” on page 53
3. “Configuring a backup pool” on page 53
4. “Labeling the device” on page 54
Note: All of the procedures in this section must be performed on a NetWorker server release
7.3.3 or later. Access the NetWorker server from the Administration window of the NetWorker
Management Console.
The EMC NetWorker Release 7.5 Service Pack 1 Multiplatform Version Administration
Guide provides more detailed information about performing the steps included here.
Configuring a device
Configure a media device for snapshot metadata. Configure a tape, file, or advanced
file type device. For best performance, configure a file or advanced file type device so
that a snapshot can be recovered without the potential delay associated with
retrieving a tape.
To configure a device:
1. In the NetWorker Administration window of the NetWorker Management
Console, select Devices on the taskbar.
2. In the expanded left pane, select Devices.
3. From the File menu, select New.
4. In the Name attribute, replace the default name with the path and name of the
device:
• If the device is configured on the NetWorker server’s storage node, the name
is the simple device path. For example, C:\tmp\d0 for a file type device.
A tape device would have a format similar to \\.\Tape0.
• If the device is configured on a remote storage node, the name must indicate
that the storage node is remote, by including rd= and the name of the remote
storage node in the device path. For example, if the remote storage node is
neptune, then the device path might be rd=neptune:c:\tmp\d0.
5. In the Comment field, specify a comment for the device.
6. From the Media Type attribute, select the appropriate media type.
7. Select the Auto Media Management attribute.
8. Click Configuration.
9. In the Target Sessions attribute, type or select a value. Set this attribute to a value
that will speed up the backup.
10. Click OK.
52
Configuring a label template
Labels identify the kind of data that is stored on the volumes in a backup pool. Label
templates define a naming convention for labels. Create a label template for volumes
that are used to contain snapshot metadata.
To configure a label template:
1. In the Administration window of the NetWorker Management Console, click
Media.
2. In the expanded left pane, select Label Templates.
4. In the Name attribute, type a name that identifies the series of labels as belonging
to volumes for snapshot metadata.
5. In the Comment attribute, specify a comment for the label.
6. In the Fields attribute, specify a text name such as Snapshot_Metadata and on a
separate line, specify a numeric range such as 001-999 or a text range such as
aa-zz.
These attributes are used to incrementally identify each label.
7. Click OK.
Configuring a backup pool
To configure a backup pool to store snapshot metadata:
Media.
2. In the expanded left pane, select Media Pools.
4. For the Name attribute, type a name that matches the label template.
5. In the Comment attribute, specify a comment for the pool.
6. Select the Enabled attribute.
7. For the Pool type attribute, select the backup pool type.
8. For the Label template attribute, select the matching label template.
9. Do not select a group in the Groups attribute.
Do not associate the pool with a particular group. This pool may be associated
with several backup groups, so a single group should not be specified here.
10. Click the Selection Criteria tab.
11. For the Clients attribute, specify a NMM Client. A value must be entered for this
attribute because no group was specified.
12. In the Devices attribute, select each device that can accept snapshot rollovers for
this pool.
13. Click OK.
Task 1: Configure a backup pool for snapshot operations
53
Labeling the device
Before a device can be used for backup, it must be labeled.
To label a device:
Devices.
2. In the right pane, right-click on the name of the device and select Label.
3. In the Pool attribute, select the backup pool that was created.
4. Select the Mount after Labeling attribute.
5. Click OK.
54
Task 2: Configure snapshot policies
A snapshot policy determines the following:
◆
When and how often snapshots are created
◆
Number of snapshots, and if any are retained as persistent snapshots
◆
Which snapshots, if any, are rolled over to conventional storage
◆
When snapshots expire and become eligible for deletion
Note: Snapshots can also be manually rolled over or deleted. Chapter 14, “Snapshot
Management,” describes these procedures in more detail.
You can specify a preconfigured policy or create a custom snapshot policy. The
following preconfigured snapshot policies exist:
◆
Serverless
A single snapshot is taken per day. The data is then rolled over to conventional
storage and the snapshot is deleted.
◆
Daily
Eight snapshots are taken per day. The data in the first snapshot is rolled over to
conventional storage. Each snapshot has 24-hour expiration policy.
CAUTION
!
To protect against a disaster, ensure that all volumes and Windows system
components are backed up. Also ensure that the associated snapshot policy
includes at least one rollover to a conventional backup. The snapshot policy
for volumes and Windows system components should specify in the Backup
Snapshots field a value of “first,” “last,” or “all.” Do not specify “none.”
Disaster recovery can only be performed from a conventional backup.
Snapshot policies for SYSTEM COMPONENTS backup
To back up SYSTEM COMPONENTS, the snapshot policy must specify “all.” NMM
enforces this policy to ensure that a permanent backup of the snapshot exists.
Attempting to restore system components from a persistent snapshot may invalidate
the snapshot and leave the system in an unrecoverable state.
Creating a snapshot policy
To create a custom snapshot policy:
Configuration.
2. In the expanded left pane, select Snapshot Policies.
4. In the Name attribute, type a name for the snapshot policy.
5. In the Comment attribute, type a comment for the snapshot policy.
55
6. In the Number of Snapshots attribute, enter the number of snapshots to be
created per day.
The Number of Snapshots attribute works in conjunction with the Group resource
Interval and Start Time attributes. The value for the Start Time and Interval
attributes must allow the specified number of snapshots to be created in a 24-hour
period. For example, to take four snapshots per day and the first snapshot must
occur at 4:00 A. M. (Start Time attribute), the Interval attribute must be set to 5
hours or less:
Number of Snapshots < (24:00 - Start Time) / Interval
“Task 4: Configure a backup group” on page 59 provides additional information
about Snapshots and policies.
7. In the Retain Snapshots attribute, enter the maximum number of snapshots that
can be retained for a specified period of time before being expired. The period of
time is specified in the Snapshot Expiration Policy attribute.
Note: Use a minimum Retain Snapshot value of 2. This will insure that the most recent
snapshot is retained while the new snapshot is created. A Retain Snapshot value of 1 will
delete the only snapshot copy while creating the new snapshot.
8. In the Snapshots Expiration Policy attribute, select a preconfigured expiration
policy to determine how long snapshots can be retained before being deleted.
9. In the Backup Snapshots attribute, specify which snapshots will be rolled over to
a conventional storage medium.
Valid values are All, None, First, Last, Every n.
!
CAUTION
components are backed up. Also, ensure that the associated snapshot policy
for volumes and Windows system components should specify in the Backup
Snapshots field a value of “first,” “last,” or “all.” Do not specify “none.”
Disaster recovery can only be performed from a conventional backup.
If the Backup Snapshot attribute is set to None, you can still roll over the snapshot
manually. Chapter 14, “Snapshot Management,” provides additional information
about snapshots.
10. Click OK.
Note: For Data Protection Manager backups the Backup Snapshot attribute must be set to
All. Backup will fail if any other value is used.
56
Snapshot policy examples
Table 6 on page 57 depicts settings for a conventional backup. Four snapshots per
day are taken. All snapshots are rolled over to a conventional storage medium. Each
snapshot is deleted after it is rolled over.
Table 6
Conventional backup settings
Name
Conventional backup
Comment
Rollover and then delete
Number of snapshots
4
Retain snapshots
0
Snapshot expiration policy
Day
Backup snapshots
All
Table 7 on page 57 depicts settings for an instant backup. Four snapshots per day are
taken. Up to 28 snapshots are retained for a maximum period of one week. No
snapshots are rolled over.
Table 7
Instant snapshot policy
Name
Instant backup
Comment
Persistent snapshots with no rollovers
Number of snapshots
4
Retain snapshots
28
Week
Backup snapshots
None
Table 8 on page 57 depicts settings for an instant backup with rollover to a
conventional medium. Four snapshots per day are taken. Up to 28 snapshots are
retained for a maximum period of one week. Every first and third snapshots are
rolled over.
Table 8
Instant snapshot with rollover policy
Name
Instant and conventional backup
Comment
Persistent snapshots with rollover to tape
Number of snapshots
4
Retain snapshots
28
Week
Backup snapshots
Every 1
Every 3
57
Task 3: Configure a backup schedule
Backup schedules determine the days on which full or incremental backups are run.
Full backups include all of the data specified in an NMM Client resource save set,
while incremental backups include only the data that has changed since the last
backup.
Table 9 on page 58 contains the considerations for configuring a backup schedule for
an NMM Client. The procedure for creating backup schedules for an NMM Client is
the same as for creating a backup schedule for a regular NetWorker client.
Table 9
Considerations for NMM Client backup schedules
Backup levels
Consideration
Full level
These backups are supported for all types of data.
Incremental
backups
These are supported only for:
• Backup of volume directories or paths such as E:\ or
E:\business_files\.
• Backup of Microsoft Exchange Server 2007.
Note: Incremental backup of Exchange Server 2007 requires
NetWorker 7.5.1, or NetWorker 7.4.
Level 1 to 9
backups
These are supported only for backup of volume directories or paths
such as E:\ or E:\business_files\.
Note: The information in Table 9 on page 58 applies only to rollover backup operations.
Snapshots are always taken at full level.
58
Task 4: Configure a backup group
Backup groups enable you to balance backup loads to reduce the impact on your
storage and network resources. Use a backup group to assign the following attributes
to a set of Client resources:
◆
Snapshot policy
◆
Backup start times
◆
Schedules
◆
Backup pools
Note: A Client resource or backup group cannot specify volumes that use both CLARiiON and
Symmetrix storage systems. For hosts that have both CLARiiON and Symmetrix volumes,
create separate backup groups and Client resources. Each backup group and its Client
resources can specify volumes from only one type of storage system.
To configure a backup group:
Configuration.
2. In the expanded left pane, select Groups.
4. In the Name attribute, type a name for the backup group.
5. In the Comment attribute, type a description.
6. For the Start Time attribute, enter the time when the first snapshot is to be
created.
7. For the Autostart attribute, select Enabled.
8. Click the Snapshot attribute so that a check mark appears beside it.
9. For the Snapshot Policy attribute, select a snapshot policy. “Task 2: Configure
snapshot policies” on page 55 provides more information about snapshot
policies.
10. For the Snapshot Pool attribute, select a pool that was created for the snapshot.
“Task 1: Configure a backup pool for snapshot operations” on page 52 provides
more information about pools.
11. Click the Advanced tab.
12. For the Interval attribute, specify how often a snapshot will be created.
The Interval attribute works in conjunction with the backup group Start Time
attribute and the snapshot policy Number of Snapshots attribute. For example, if
you want to take four snapshots per day (Number of Snapshots attribute) and the
first snapshot must occur at 4:00 A.M. (Start Time attribute), then the Interval
attribute must be set to 5 hours or less:
Number of Snapshots < (24:00 - Start Time) / Interval
13. Ensure that Restart Window attribute value is less than or equal to the Interval
attribute value.
14. Set the Client Retries attribute to 0 (zero).
15. Click OK to create the backup group.
59
!
CAUTION
Once you have started a snapshot of a save group, do not interrupt or halt the
snapshot process. For example, in an Exchange Server backup, the
nsrsnap_vss_save.exe process on the production server and the eseutil process on
the proxy may continue to run after the snapshot is halted. Any attempt to stop a
group in NetWorker Management Console will take a long time to complete.
60
Task 5: Configure a Client resource
A Client resource specifies what to include in a snapshot of an NMM Client. Client
resources are associated with other backup resources such as groups and snapshot
policies.
You can create multiple Client resources for the same NMM Client host. In this way,
you can apply different backup attributes to different types of information on the
same host. For example, if the NMM Client host is an Exchange Server, you can create
one Client resource to back up the Exchange databases and create another Client
resource to back up Windows system component data. This way, you can back up
Exchange databases many times a day, and back up Windows system component
data once a day.
The general process for configuring a Client resource is the same for all applications
or systems. The following list describes some of the differences:
◆
System volumes and components use save sets of All, volume name, mount
points, or the save set name SYSTEM COMPONENTS.
◆
System volumes and components do not use application variables.
◆
Applications use the save set APPLICATIONS:\Name of writer.
◆
Some applications backed up by NMM have settings and requirements specific to
the application. These settings are specified by application variables, described in
the backup sections for each application.
Configuring a Client resource for System volumes, components, and applications
The Client resource configuration task is detailed in separate sections for the
following Systems volumes, components, and applications:
◆
“Configuring a Client Resource for deduplication” on page 78
◆
“Performing system backups” on page 84
◆
◆
◆
◆
◆
“Performing SharePoint 2007 granular backups” on page 175
◆
◆
◆
Specifying save sets for application data
Table 10 on page 62 lists the save set syntax to specify for supported types of
application data. Specify application data save sets in the Save Set attribute of a Client
resource.
“Displaying valid application data save sets” on page 63 describes how to find the
exact name of a particular Exchange storage group or SQL Server instance and
database combination. save set syntax.
61
Table 10
Save set syntax
Type of backup data
Save set syntax
DFS data (Windows Server 2003 only)
APPLICATIONS:\DFS Replication service
writer
Also back up the SYSTEM COMPONENTS:\ save set on
the namespace server (usually the domain controller) that is
associated with the DFS configuration.
Note: In Windows Server 2008, DFS Replication service writer is
backed up as part of SYSTEM COMPONENTS:\.
62
ADAM data
APPLICATIONS:\ADAM (instance_name)
Writer
For example, the ADAM instance named AddressBook would be
typed as:
APPLICATIONS:\ADAM (AddressBook) Writer
Chapter 13, “Microsoft Active Directory Backup and Recovery,”
provides more information about Active Directory services.
In Windows Server 2008, Active Directory Lightweight Directory
Services (AD LDS) includes the Active Directory Application
Mode (ADAM) functionality provided in Windows Server 2003.
Active Directory
Chapter 13, “Microsoft Active Directory Backup and Recovery,”
provides more information about Active Directory services.
“Displaying valid Exchange data save sets” on page 198
provides more information about Exchange save sets and
syntax.
“Displaying valid SQL data save sets” on page 116 provides
more information about Microsoft SQL Server save sets and
syntax.
Microsoft System Center Data Protection
Manager 2007
“Displaying valid DPM data save sets” on page 252 provides
more information about DPM save sets and syntax.
Microsoft Office SharePoint Server 2007 and
Microsoft SharePoint Portal Server 2003
“Displaying valid SharePoint 2007 data save sets” on page 137
provides more information about Microsoft Office SharePoint
Server 2007.
“Displaying valid SQL data save sets for SharePoint 2003
backup” on page 154 provides more information about Microsoft
SharePoint Portal Server 2007.
Granular backup
“Displaying valid SharePoint 2007 data save sets” on page 137
provides more information about Microsoft Office SharePoint
Server 2007 granular backup.
Hyper-V
“Displaying valid Hyper-V save sets” on page 282 provides more
information about Hyper-V save sets and syntax.
Displaying valid application data save sets
To display a list of the application data save sets that are available for backup:
1. Open a command prompt on the application server.
2. From the command prompt, type the following command to list the valid
application data save set names.
If the application server is not on a virtual host, type:
nsrsnap_vss_save -?
If the application server is on a virtual host, type:
nsrsnap_vss_save -? -c virtual_server_name
where virtual_server_name is the name of the virtual host.
nsrsnap_vss_save -? -c
Note: If the application server is on a virtual host, run this command from the physical
node that is currently hosting the application server.
3. Press Enter.
The application data save sets that are available for backup are listed in a format
similar to the following:
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\master"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\model"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\msdb"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\NADATA"
Each line of output corresponds to a save set entry that you can add to the Save
Set attribute of a Client resource. Each entry that you add to the Save Set attribute
must be typed on a separate line.
URL encoding for SQL and Exchange save sets
When specifying save set names in the Save Set attribute of the Client resource, there
are cases where special characters, such as the backslash (\), must be specified by
their URL-encoded values. URL-encoded values are used to represent special
characters in the following cases:
◆
SQL Server named instances
◆
SQL Server database names
◆
Exchange storage groups
Consider the following example of a SQL Server named instance. The name of a SQL
Server named instance consists of two parts: the hostname and a unique name. These
two parts are always separated with a backslash (\) character. When specifying a SQL
Server named instance in the Save Set attribute, replace the backslash with its
URL-encoded value, which is %5C. In the following example, a SQL Server named
instance that is identified as:
CORPDBSERV1\YUKON
would be represented in the Save Set attribute as:
CORPDBSERV1%5CYUKON
63
To back up only the database named NADATA in the SQL Server named instance
CORPDBSERV1\YUKON, the complete entry in the Save Set attribute would be:
APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\NADATA
If the database name in the previous example was NA#DATA, the save set entry
would be:
APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\NA%23DATA
Table 11 on page 64 lists the most commonly used special characters and their
URLvalues.
Table 11
64
Special characters and their URL-encoded values
Special character
URL-encoded value
Special character
URL-encoded value
\
%5C
?
%3F
/
%2F
]
%5D
"
%22
[
%5B
%
%25
}
%7D
#
%23
{
%7B
&
%26
^
%5E
<
%3C
‘
%60
>
%3E
|
%7C
Application information variable settings
Table 12 on page 65 lists the variables that can be specified in the Application
information attribute of the Client resource.
Table 12
Attribute name
Description
Values
NSR_SNAP_TYPE=value
Specifies the snapshot service
provider name.
vss
This vss value is required.
NSR_DATA_MOVER=value
Specifies the hostname of the NMM
Client that moves snapshots from
the NMM Client to primary and
secondary (conventional) storage.
The host can be either a local host
or a proxy client host.
• The local host
• The proxy client hostname
If no value is entered, the local host is
used as the data mover.
If you are setting up a proxy client for a
virtual cluster server, you must also type
the proxy client hostname in the Remote
Access attribute of the Client resource.
For serverless backups, specify a proxy
client.
If setting up an Exchange Client
resource, it is recommended that you use
a proxy client. If a proxy client is
specified, Exchange consistency checks
are performed on the proxy client.
Related messages are written to the
Replication Manager log files
(erm_clientxx.log) on the proxy client.
If no proxy client is specified for an
Exchange Client resource, consistency
checks are performed on the Client
resource host. Related messages are
logged on the Client resource host in
both the Replication Manager log files
and the nmm.raw file.
In addition to the attributes listed in Table 12 on page 65, there are also
application-specific attributes:
◆
“SQL Server application information variable settings” on page 118
◆
“SharePoint 2007 application information variable settings” on page 139
◆
“SQL application information variable settings in SharePoint 2003 backup” on
page 156
◆
“Exchange application information variable settings” on page 200
◆
“DPM application information variable settings” on page 254
◆
“Hyper-V application information variable settings” on page 283
◆
“Windows Cluster application information variable settings” on page 297
65
Task 6: Configure privileges
The NMM Client must be granted NetWorker Administrator privileges in order to
perform media database operations during snapshot deletion. If you set up a proxy
client for the NMM Client, grant the proxy client NetWorker Administrator
privileges. If you are setting up a NMM Client in a cluster, grant NetWorker
administrator privileges to each cluster node and virtual server.
To add configure privileges:
Configuration.
2. In the expanded left pane, click User Groups.
3. Right-click the Administrators group, then select Properties.
The Properties dialog box appears.
4. In the Users attribute, add the following values for the NMM Client host. Place
each value on a separate line:
user=administrator,host=NMM_client_host
user=system,host=NMM_client_host
where NMM_client_host is the DNS hostname of the NMM Client.
If a proxy client is set up for the NMM Client, add the following values:
user=administrator,host=NMM_clientproxy_host
user=system,host=NMM_clientproxy_host
If the NMM Client is installed in a cluster, grant NetWorker Administrator
privileges to each cluster node and virtual server.
5. Click OK.
66
Task 7: Configure a proxy client
A proxy client is a separate host that acts as a data mover for a NMM Client. By
default, the NMM Client acts as its own data mover. However, a proxy client is
required for serverless backups. Serverless backups free the NMM Client from much
of the processing involved in a snapshot backup. In a serverless backup, a snapshot is
created on the NMM Client and then immediately moved to the proxy client for
further processing.
A proxy client can also be set up on a NetWorker storage node. In this way network
traffic that is generated when a snapshot is rolled over to a conventional backup
medium is avoided.
Table 13 on page 67 lists the requirements for setting up a proxy client.
Table 13
Configuration requirements for a proxy client (page 1 of 2)
Requirement
Description
General requirements
Installation
The proxy client must have the same version of the NMM Client software, and
operating system version and patch level, as the host for which it is acting as
the proxy client. The NetWorker Module for Microsoft Applications Release 2.2
Installation Guide provides more information about installation.
Configuration
Create a NMM Client resource for the proxy client host. “Task 5: Configure a
Client resource” on page 61 provides more information about performing this
task.
Type the proxy client host name in the Application Information and the Remote
Access attributes of the NMM Client resource that will use the services of the
proxy client.
Connectivity
The proxy client must have connectivity between the NMM Client, the
NetWorker server, and the storage subsystem on which snapshots reside.
Serverless
backups
A proxy client is required for serverless backups.
Additionally, serverless backups are supported only with VSS hardware
providers and depend on the capabilities of the storage subsystem. For the
EMC VSS Provider, serverless backups are supported for both CLARiiON and
Symmetrix storage systems.
Rollback recovery
The NMM Client to which the snapshot is to be rolled back must be supported
with a proxy client. “Roll back a snapshot” on page 327 provides more
information rollback requirements.
CLARiiON
For proxy clients using CLARiiON storage and the EMC VSS Provider, at least
one CLARiiON LUN must be permanently assigned to the proxy host.
Third-party
hardware
providers
Hardware providers may have additional configuration requirements for proxy
clients. Refer to the vendor’s product documentation for details on these
requirements.
Task 7: Configure a proxy client
67
Table 13
Configuration requirements for a proxy client (page 2 of 2)
Requirement
Description
Additional requirements for an Exchange Server
68
Exchange Server
performance
Set up a proxy client for an Exchange Server. A proxy client performs the
processing associated with the Exchange Server consistency checker utility,
eseutil.exe, thus reducing the load on the Exchange Server.
Consistency
checker utility
The Exchange Server consistency checker utility, eseutil.exe, must be
installed on the proxy client. The version of eseutil.exe must be of the same
version (service pack level) as the Exchange Server. For example, if
Exchange Server 2003 Service Pack 2 is installed on the Exchange Server,
then the version of eseutil.exe that is installed on the proxy client must be
from the Exchange Server 2003 SP2 Management utilities. Failure to do so
will cause the consistency checker utility to report errors even when the
databases are valid.
Install the eseutil.exe utility from the Exchange Management Tools located on
the Exchange Server CD.
Installation path to
consistency
checker utility
If the installation of the Exchange consistency checker utility, eseutil.exe, is
not performed from the Exchange Server installation program, ensure that the
path to the eseutil.exe program is in the $PATH environment variable.
Exchange mount
paths
Use the NSR_ALT_PATH variable to specify a path to mount snapshots on the
proxy client. “Application information variable settings” on page 65 provides
more information about these settings.
Backing up a clustered NMM Client
To configure a virtual NMM Client for backup, complete all of the tasks required to
back up a regular NMM Client. However, there are some considerations to be aware
of when setting up a virtual NMM Client, which are outlined in Table 14 on page 69.
Table 14
Additional steps and considerations for backing up a clustered Client resource
Backup task
Consideration
“Task 5: Configure a Client
resource” on page 61
Configure a Client resource for each virtual server that is being backed up and
each physical node in the cluster on which the virtual server may execute.
Type the names of the physical nodes of the cluster in the Remote Access
attribute of each virtual Client resource.
“Task 6: Configure privileges” on
page 66
Set up NetWorker administrator privileges for each physical node and proxy
client in the cluster.
“Task 7: Configure a proxy client”
on page 67
If a hardware provider such as the EMC VSS Provider is being used, a proxy
client must be configured for a clustered NMM Client.
NetWorker pathownerignore cluster functionality is not supported
The pathownerignore cluster functionality is not supported in NMM. This restriction
is not enforced by NMM code. In some circumstances the path owner may be ignored
by NMM so that the data from a clustered disk is backed up under the indices of the
physical node. But recovers of that data will fail.
Clustered disks must be backed up under a virtual cluster client. Do not attempt to
use the pathownerignore functionality. Ensure that a nsr\bin\pathownerignore file
does not exist.
Cluster failover and backups
If a node within a cluster undergoes failover during a backup operation, the
operation will fail. The next scheduled backup operation will be the next valid
backup.
If you are using a hardware provider such as the EMC VSS Provider, use the disk
management utilities provided with the associated hardware to delete any resources
that may be left in an indeterminate state as a result of the failed backup. For
example, a CLARiiON storage solution may have an InActive snapshot as the result
of a failed backup. In this case, use the EMC Navisphere® user interface or the
NaviCLI command line interface to search for and delete the inactive snapshot.
Backing up a clustered NMM Client
69
Moving an NMM Client to another NetWorker server
Do not set up scheduled backups for an NMM Client on multiple NetWorker servers.
An NMM Client should be protected by only one NetWorker server.
To move a NMM Client to a different NetWorker server:
1. On the NetWorker server that you are moving from, disable or delete the Client
resources that are set up for the NMM Client.
You can disable a Client resource for scheduled backup by clearing the Scheduled
backup attribute in the Client resource.
2. On the NMM Client, open the NMM Client software and manually delete any
snapshots for the client. “Delete a snapshot” on page 330 provides more
information about deleting snapshots manually.
Before deleting the snapshots, you can manually roll them over to a conventional
backup. In this way, the data in the snapshot can be recovered.
3. On the NetWorker server that you are moving to, set up scheduled backups for
the NMM Client. “Backup configuration roadmap” on page 48 provides more
information about setting up scheduled backups.
70
3
Deduplication Backup
and Recovery
◆
◆
◆
◆
◆
Data deduplication........................................................................................................
Avamar and NetWorker server configuration ..........................................................
Deduplication backups.................................................................................................
Configuring a Client Resource for deduplication ....................................................
Recovering deduplicated data.....................................................................................
72
75
76
78
80
71
Data deduplication
This chapter introduces and discusses the Avamar online data protection solution
integrated in NetWorker and NetWorker Module for Microsoft Applications (NMM).
NetWorker Deduplication, which leverages Avamar Data Stores, solves the
challenges associated with traditional backup, enabling fast, reliable backup and
recovery for remote offices, and data center LANs. Avamar is backup and recovery
software that utilizes patented global data deduplication technology to identify
redundant sub-file data segments at the source, reducing daily backup data by up to
500x — before it is transferred across the network and stored to disk. This allows
companies to perform daily full backups even across congested networks and limited
WAN links.
Key NetWorker Deduplication differentiators are:
72
◆
Deduplicates backup data at the source—before transfer across the network
◆
Enables fast, daily full backups across existing networks and infrastructure
◆
Reduces required daily network bandwidth by up to 500x
◆
Up to 10x faster backups
◆
Data encrypted in flight and at rest
◆
Patented RAIN technology provides fault tolerance across nodes and eliminates
single points of failure
◆
Scalable grid architecture
◆
Global data deduplication reduces total backup storage by up to 50x
◆
Recoverability verified daily—no surprises
◆
Centralized web-based management
◆
Simple one-step recovery
◆
Flexible deployment options, including EMC Avamar Data Store package
Data deduplication replaces multiple copies of the same data scattered across
networks with a reference to a single instance or source for the data. Edited files also
have tremendous redundancy with previous versions.
Avamar uses patented global data deduplication and single-instance storage
technology for more efficient backup operations and storage:
◆
Backup data size is reduced at the source—before it is transferred across the
network and stored to disk.
◆
Redundancy is eliminated at both the file and the subfile data segment level.
As a result Avamar delivers fast and efficient daily full backups, despite slow or
congested infrastructure, and data can be encrypted for added security.
Related information about Avamar server and NetWorker server
Extensive documentation for the Avamar server setup and configuration is available
for download at http://Powerlink.EMC.com. In addition to the Avamar product
documentation, the following documents provide more information:
◆
The EMC white paper Efficient Data Protection with EMC Avamar Global
De-duplication Software provides a technical overview of EMC Avamar backup
and recovery software.
◆
The EMC NetWorker Release 7.5 Administration Guide describes creating a
deduplication storage node, replication nodes, and other NetWorker
deduplication features and services.
Avamar integration in NMM
The integration of Avamar in NMM provides deduplication for Microsoft
Applications through the familiar NetWorker features and GUI. This integration
provides the ability to:
◆
Configure deduplication for an NMM client resource in NetWorker Management
Console (NMC).
The deduplication backup settings are part of the Client Resource configuration
GUI. It is as simple as selecting a checkbox to enable deduplication, and then
selecting the Deduplication server.
◆
Perform deduplication backup at snapshot rollover time.
When deduplication is selected in the Client Resource, NMM automatically
performs deduplication during the backup—no separate or additional user input
is required to perform deduplication.
◆
Perform deduplication recover from rollover.
The process for recovering data from a deduplication node is basically the same
as that for recovering from a storage node.
◆
File and data sizes vary between applications, so Avamar optimizes compression
by analyzing the data in the save set and using the best chunking size for that
data.
Data deduplication
73
Windows OS, application, and configuration support
The EMC Information Protection Software Compatibility Guide, available on the EMC
Powerlink website http://Powerlink.EMC.com, contains additional and the most
up-to-date information about NMM compatibility. NMM supports deduplication on
the following:
Operating systems
◆
Windows Server 2008 (x86, x64)
◆
Windows Server 2003 R2 (x86, x64)
◆
Windows Server 2003 (x86, x64)
Applications
◆
Microsoft SQL Server 2008 x86/x64
◆
Microsoft SQL Server 2005 x86/x64
◆
Microsoft Office SharePoint Server 2007 (VSS, optimized, and granular) x86/x64
◆
Microsoft SharePoint Server 2003 x86/x64
◆
Microsoft Exchange Server 2007 x64
◆
Microsoft Exchange Server 2003 x86
◆
Windows Server 2008 with Hyper-V
◆
Windows Server 2008, Server Core Installation, with Hyper-V
◆
System Components and File System
◆
◆
Configurations
◆
System Provider
◆
Hardware Provider: Proxy Backup Configurations
◆
Cluster configurations are supported for the operating systems, applications, and
configurations in the preceding lists.
Unsupported for deduplication
NMM does not support deduplication for the following applications, features, and
configurations:
◆
Applications:
• Active Directory granular
• Microsoft System Center Data Protection Manager 2007
◆
NetWorker features:
• Cloning/Staging
• AES Encryption
• Compression
◆
74
Windows configurations: EFS File Systems
Avamar and NetWorker server configuration
Before a NMM backup client can be configured to use Avamar deduplication, the
Avamar server and NetWorker server must be installed to receive deduplication data
and process backups. For the NetWorker server, NetWorker Release 7.5 SP1or later is
required. For the Avamar server hardware, Avamar Axiom v4.0.2 or later is required.
Configuring Avamar and the NetWorker server to receive deduplication data and process
backups
To configure Avamar and the NetWorker server to receive deduplication data and
process backups:
1. Set up and configure the Avamar server.
The Avamar hardware and software documentation provides information on
Avamar setup and configuration.
2. Create deduplication nodes and replication nodes on the Avamar server.
Perform this task through the Avamar user interface and Avamar documentation.
3. Configure deduplication node and replication node settings on the Networker
server.
Configure the deduplication node through the NetWorker server’s
Administration interface. The EMC NetWorker Release 7.5 Administration Guide
provides more information about configuring the deduplication node.
4. Create a replication node for a deduplication node, if required by the NetWorker
deduplication node for automatic failover of data backup nodes.
The EMC NetWorker Release 7.5 Administration Guide provides more information
about configuring the replication node.
Setting DNS resolution for NetWorker Deduplication Node and NMM
The NetWorker Deduplication Node and the NMM hosts (one or more of the
following: NMM client, NMM proxy client, or NMM proxy storage node) need to
have DNS resolution between nodes. DNS resolution must be working prior to
starting deduplication backups:
◆
On the NMM hosts, make sure the hosts have DNS name resolution for the
NetWorker Deduplication Node.
If the NMM hosts do not have DNS resolution for the NetWorker Deduplication
Node, then that needs to be repaired. Check your operating system
documentation for information on how to fix DNS resolution problems.
◆
On the NetWorker Deduplication Node, make sure it has DNS name resolution to
the NMM hosts.
If the NetWorker Deduplication Node does not have DNS resolution for the
NMM hosts, then that needs to be repaired. Check your operating system
documentation for information on how to fix DNS resolution problems.
Avamar and NetWorker server configuration
75
Deduplication backups
All typical client operations, such as adding, deleting, and drag-and-drop, work the
same for deduplication clients as they do for other types of clients. A deduplication
save set is treated as a regular NetWorker save set, although it actually has two parts:
metadata (hash info) and the backed-up client data. Only the metadata is stored on a
NetWorker storage node. The backed-up client data is stored on a deduplication node
(Avamar server). The metadata can be cloned in the usual manner, while a copy of the
backed-up data can be replicated on another deduplication node.
When configuring a NMM Client Resource for deduplication, there are several
special requirements or considerations. The following topics provide more
information:
◆
“Backup levels and schedules for deduplication clients” on page 76
◆
“Retention policies” on page 76
◆
“Backup configuration notes” on page 77
Backup levels and schedules for deduplication clients
When a NMM deduplication client is backed up for the first time after deduplication
has been enabled, a full backup of the client data must be performed. Subsequent
backups of the client then take advantage of deduplication to back up only those data
segments that have changed since the previous backup. The exception to this is that
full deduplication backups always include a complete backup of the index.
Note: If the deduplication node for an existing deduplication client is changed or if another
type of client is upgraded to a deduplication client, a new full backup of the client data must be
performed, regardless whether the most recent backup performed was a full backup.
The initial full backup of a deduplication client takes longer than a regular backup of
the same client data due to the initial overhead involved in deduplicating the data.
All subsequent deduplication backups benefit from the fact that now only the data
segments (not files) that have changed since the previous backup will be backed up.
The choice of subsequent backup levels, however, depends on what is more
important to the user: backup performance or recovery performance. Deduplication
clients should use backup schedules that support that priority.
Deduplication backups must be scheduled to avoid the deduplication node's
read-only periods, when such cron jobs as checkpoint and garbage-collection are run.
Refer to the Avamar server documentation for how to do this.
Retention policies
If a volume contains one or more deduplication save sets, the resource for the
deduplication node that was used to create the backup must exist when the save sets
pass their retention policy. If the resource for the deduplication node has been
deleted, the volume cannot be made recyclable or relabeled. Furthermore, when
deduplication save sets pass their retention time, the NetWorker server will begin the
process of deleting the deduplicated data from the deduplication node. Therefore,
deduplication data may not be recoverable using the scanner program once the
deduplication save set has passed its retention time.
The EMC NetWorker Release 7.5 Administration Guide provides more information about
Retention Policies, and the save set recover and scanner programs.
76
Backup configuration notes
The following notes are described in more detail in the EMC NetWorker Release 7.5
Administration Guide:
◆
Do not place both regular and deduplication clients in the same groups.
◆
Insure all deduplication clients in the same group have the same deduplication
node setting.
◆
Directives cannot be applied if deduplication is used for a client.
◆
Media pools do not apply when deduplication is used.
◆
The cloning of deduplication backups is handled differently from that of regular
backups. Only the metadata is stored on a NetWorker storage node, while the
backed up data is stored on a deduplication node.
An alternate method is available to handle tape backups of the deduplication
client data. Create a second instance of the client to be backed up, but do not
configure the second instance as a deduplication client. If using multiple backup
schedules, then the backup level must be full for all groups. This is required
because the backup level and the last backup time are calculated using both the
group name and save set name, and the groups would interfere with each other.
Deduplication backups
77
Configuring a Client Resource for deduplication
When NMM is installed, it automatically includes support for deduplication. You do
not need to install anything extra on the NMM client. After the Avamar server and
NetWorker server have been configured for deduplication, configure a Client
Resource to use deduplication.
Configuring a deduplication client
After creating a deduplication node, configure a NetWorker deduplication client.
There are two workflows you can use:
◆
Create a new Client Resource, enable deduplication, and then configure the Client
Resource with the application-specific settings for whichever Microsoft
application you want to back up.
◆
Edit an existing Client Resource, for example a NMM Client Resource you
previously created for Microsoft SQL Server backup.
◆
You can use this procedure for configuring deduplication for any client resource
listed in “Windows OS, application, and configuration support” on page 74,
except for configuring deduplication for an Exchange client running on a passive
node in a CCR environment, which is described in “Configuring deduplication
backup for a passive node in a CCR environment” on page 218:
1. On the NetWorker server, open the NetWorker Management Console, open the
NetWorker Administration window, and then click the Configuration button.
2. Select Clients in the navigation tree. The Clients table appears. It lists any clients
that have already been created:
• To add a new Client Resource, right-click Clients in the navigation tree or
right-click any client in the Clients table, and then select New. The Create
Client window is displayed.
• To edit an existing Client Resource, right-click the client in the Clients table,
and then select Properties. The Properties window is displayed.
3. On the General tab, enter the information as for any client, with the following
exceptions:
• The Name of the client must be a computer qualified to be a NetWorker
deduplication client: it must run an operating system that supports
deduplication. The Name must be a fully qualified domain name (FQDN).
• In the Backup area:
– Ensure that the deduplication client has been assigned to a group that
contains only deduplication clients. Do not mix regular and de-duplication
clients within a group.
See “Task 4: Configure a backup group” on page 59 for information about
creating backup groups.
– A recommended backup schedule for deduplication clients depends on
whether your priority is faster backups or faster, less complicated
recoveries.
See “Task 3: Configure a backup schedule” on page 58 for general
information about setting backup schedules. See “Backup levels and
schedules for deduplication clients” on page 76 for backup level and
schedule information that is specific to deduplication clients.
78
Note: For deduplication clients, the Parallelism attribute in the Client resource should be
set to less than or equal to 4.
4. Click the Apps & Modules tab.
5. In the Access area, leave the Remote user and Password fields empty.
6. In the De-duplication area, select the De-duplication backup attribute to enable
this client for deduplication backups.
7. Select the name of the de-duplication node to which this client’s backup data will
be sent. This is the deduplication node created in “Configuring Avamar and the
NetWorker server to receive deduplication data and process backups” on
page 75.
Note: All clients in a group must have the same deduplication node value.
This step links this client with its own deduplication node. Do not select the name
of a replication node here.
!
IMPORTANT
If the de-duplication node for this client’s backup data changes, the next backup
done must be a level 0 (full) backup.
8. Complete the remaining configuration information as for any regular client.
“Configuring a Client resource for System volumes, components, and applications”
on page 61 lists the application-specific Client resource configuration procedures for
each System or application backup.
Configuring a Client Resource for deduplication
79
Recovering deduplicated data
The process for recovering data from a deduplication node is basically the same as
that for recovering from a storage node. However, there is an underlying difference in
where the information is kept. Only the metadata, which is hash information, is
stored (or cloned or staged) on a NetWorker storage node. The backed-up data from a
deduplication client is stored on a deduplication node that is an Avamar server. Both
the deduplication node and the volume on the recovered side must be on line during
the recovery of deduplicated data.
Note: Under some conditions, if recovery from the primary deduplication node fails,
autorecovery will use a replication node. The EMC NetWorker Release 7.5 Administration Guide
provides detailed information about recovery from a replication node.
Deleting deduplication save sets
The process for deleting deduplication save sets is performed on the NetWorker and
Avamar servers, and is described in the EMC NetWorker Release 7.5 Administration Guide.
Querying deduplication save sets using mminfo
You can use the mminfo program to query information about save sets created using
deduplication. The output of mminfo identifies save sets that were created with
deduplication by using the dedupe flag. To limit the output of mminfo to only those
save sets created using deduplication, use the -q dedupe option. For example, to
query deduplication save sets for the local host, use the following command:
mminfo -S -q dedupe
Be aware that this command does not list the following:
◆
Empty save sets
◆
Save sets in which nothing was backed up as the result of skip directives
The previous save sets are treated as regular save sets, not as deduplicated save sets.
Additional NetWorker server deduplication operation information
Additional deduplication operation information is available in the following
NetWorker features:
◆
Backup statistics basic reports
◆
Events
The EMC NetWorker Release 7.5 Administration Guide describes the deduplication
information provided by these features.
80
4
System Backup and
Recovery
◆
◆
◆
◆
◆
◆
◆
◆
◆
Windows system backup ............................................................................................. 82
Performing system backups ........................................................................................ 84
Performing a recovery .................................................................................................. 86
Recovering file system snapshots that have not been rolled over ....................... 101
Booting Windows in Directory Service Restore mode........................................... 102
Performing an authoritative recovery of NTDS or FRS data................................ 103
Performing a directed recovery................................................................................. 105
Recovering the Windows system configuration to an earlier state ..................... 108
Disaster recovery......................................................................................................... 109
81
Windows system backup
This chapter supplements the overall NetWorker Module for Microsoft Application
(NMM) backup procedures in Chapter 2, “Configuring a Scheduled Backup,” and the
recovery procedures in with the specific details needed to back up and recover
Windows System data, such as Windows volumes and system components.
Windows system support
NMM supports backup and recovery of Microsoft Windows Server 2008, 2003, and
Storage Server 2003. “Windows operating system support” on page 24 provides a
more detailed list of the versions supported.
NMM also provides backup and recovery for Active Directory and Windows clusters:
◆
Chapter 13, “Microsoft Active Directory Backup and Recovery” provides specific
details for Active Directory backup and recovery.
◆
Chapter 12, “Microsoft Windows Cluster Backup and Recovery,” provides
specific details about cluster backup and recovery of Microsoft Windows Server
2008 and Microsoft Windows Server 2003.
Specifying system save sets for recovery
You can back up system data by specifying the volume and system components.
A volume save set can be as simple as a drive letter, or it can be a path or mount
point.
The SYSTEM COMPONENTS:\ save set contains many individual system writers.
The individual system writers cannot be specified in a backup save set, they are all
backed up in the SYSTEM COMPONENTS:\save set.
The SYSTEM COMPONENTS:\ save set includes the following writers:
82
◆
Background Intelligent Transfer Service (BITS)
◆
COM+ Registration Database (REGDB)
◆
Certificate Authority
◆
Cluster Database (Windows Server 2003)
◆
Cluster Writer (Windows Server 2008)
◆
Dynamic Host Configuration Protocol (DHCP) Jet
◆
Event log
◆
File Replication Server (FRS)
◆
File Storage Resource Manager (FSRM)
◆
◆
IIS Configuration Writer (Windows Server 2008)
◆
IIS Metabase
◆
Microsoft SQL Server Desktop Engine (MSDE) for SQL 2000 embedded databases
◆
MSSearch Service Writer (Windows Server 2008)
◆
NPS VSS Writer (Network Policy and Access Service, Windows Server 2008)
◆
NTDS (Active Directory Schema)
◆
Registry
◆
Removable Storage Manager (RSM)
◆
Remote Storage Manager
◆
Shadow Copy Optimization Writer (Windows Server 2008)
◆
SQL Embedded (SQL 2005 Express)
◆
Terminal Services Gateway Writer (Windows Server 2008)
◆
Terminal Services Licensing
◆
Windows Internet Name Service (WINS)
◆
Windows Management Instrumentation (WMI)
NMM System data commands
Table 15 on page 83 describes commands NMM uses to back up and recover system
data. There are no application variables for system backups.
Table 15
NMM System data commands
Item
Description
nsrsnap_vss_save
NMM command that initiates the VSS-based backup.
nsrsnap_vss_recover
NMM command that initiates the VSS-based recovery.
Windows system backup
83
Performing system backups
The first four tasks for scheduling a backup are the same for all VSS writers that are
supported by NMM. Chapter 2, “Configuring a Scheduled Backup,” describes these
tasks in detail:
◆
“Task 1: Configure a backup pool for snapshot operations” on page 52
◆
Note: To back up “SYSTEM COMPONENTS:\” the snapshot policy must specify “all.”
NMM enforces this policy to ensure that a permanent backup of the snapshot exists.
Attempting to restore system components from a persistent snapshot may invalidate the
snapshot and leave the system in an unrecoverable state.
◆
◆
The next task is specific to configuring a Client resource for system volumes and
system components:
Task 5: Configure a System Client resource
To configure a System Client resource:
Configuration.
2. In the expanded left pane, select Clients.
4. In the Name attribute, type the hostname of the NetWorker client computer.
5. In the Comment attribute, type a description. If you are creating multiple Client
resources for the same NetWorker client host computer, use this attribute to
differentiate the purpose of each resource.
6. From the Browse Policy attribute, select a browse policy from the list. The browse
policy determines how long rolled-over data is available for quick access.
7. From the Retention Policy attribute, select a retention policy from the list. The
retention policy determines how long rolled-over data is available though not
necessarily quickly.
8. Select the Scheduled Backups attribute.
9. In the Save Set attribute, specify the components to be backed up. Place multiple
entries on separate lines:
• To specify all volume data and all available Windows system components,
type:
All
Application data such as Exchange data is not included in the save set All.
• To back up specific volumes, type the volume name. For example, to specify
the volume E and all of its subfolders and files, type:
E:\
Entries are not case-sensitive.
84
• To specify volume mount point data, place the name of the mount point on a
separate line. For example, if you have a mount point named CAD_Drawings
on volume E:\ and you want to back up all of the data on E:\ including the
data under the mount point, type the following on separate lines:
E:\
E:\CAD_Drawings
The save set All will include mount point data only if no other path to the
remote data is found during backup.
• To specify all available Windows system components such as the registration
database and Windows event log, type the save set All or type following:
SYSTEM COMPONENTS:\
Ensure that the backslash (\) character is entered with the save set name
SYSTEM COMPONENTS:\. Otherwise, backup fails and no concise error
message is provided.
!
CAUTION
components are backed up. Also ensure that the associated snapshot policy
for volumes and Windows system components should specify in the
Backup Snapshots field a value of “first,” “last,” or “all.” Do not specify
“none.” Disaster recovery can only be performed from a conventional
backup.
10. For the Group attribute, select the backup group to which this Client resource
will be added.
If Client resources for the same NMM client host are added to different backup
groups, ensure that the Start Time attribute for each backup group is spaced far
enough apart so that the backups for the host’s Client resources do not overlap.
11. For the Schedule attribute, select a backup schedule.
14. In the Backup command attribute, type the backup command. For all types of
backup data except for Active Directory conventional backups, type:
nsrsnap_vss_save.exe
“Active Directory backup and recovery” on page 308 provides additional
information about Active Directory backups.
15. In the Application information attribute, type the following variable and value:
NSR_SNAP_TYPE=vss
Type any additional variables and their values on a separate line.
16. If a proxy client is being set up for the NMM client, type the hostname of the
proxy client in the Remote Access attribute.
If the NMM client is part of a cluster, type the names of the physical nodes of the
cluster in the Remote Access attribute.
17. Click the Globals (1 of 2) tab.
Performing system backups
85
Performing a recovery
Make sure all users who will be running or configuring NetWorker Module for
Microsoft Applications (NMM) backups and recoveries have the appropriate level of
privileges for all resources they must access. Table 16 on page 86 describes the
privilege levels needed.
Table 16
Access privileges needed for NMM
Resource
Privileges needed
NetWorker Server
NetWorker Administrator or NetWorker Backup Operator
NetWorker Module for Microsoft
Applications client machine
Windows Administrator or Windows Backup Operator
All applications protected by NMM,
such as:
Windows Server
Network domain
Windows Domain Administrator
Storage node servers
Review the following options and considerations before performing the procedures
in this chapter:
◆
Chapter 13, “Microsoft Active Directory Backup and Recovery” provides more
information about recovering Active Directory (AD) data.
◆
“Recovering file system snapshots that have not been rolled over” on page 101
provides more information about recovering data from a persistent snapshot that
has not been rolled over to a conventional backup medium.
IMPORTANT
!
Attempting to restore system components from a persistent snapshot may
invalidate the snapshot and leave the system in an unrecoverable state. NMM
does not prevent an attempt to restore SYSTEM COMPONENTS from a
point-in-time backup. Chapter 14, “Snapshot Management,” describes how to
roll over a snapshot to a conventional backup.
86
◆
To recover backup data that was password-protected and encrypted with another
VSS client program, and not NetWorker VSS Client for Microsoft Windows 1.0 or
NetWorker Module for Microsoft Applications, use the recover.exe command
from the CLI (command line interface). The EMC NetWorker Command Reference
Guide, release 7.3.2 or later, provides more information about the recover.exe
command.
◆
Data that was backed up from a mount point is available for browsing at the root
of the navigation tree. For example, if the drive H:\ has a mount point named
H:\CAD_Drawings, then a separate entry for the mount point
H:\CAD_Drawings will be available at the root of the navigation tree instead of
as a subfolder under H:\.
◆
To recover data to a mount point that has been deleted, manually re-create the
mount point before recovering the data. Otherwise, the data will be recovered to a
local directory whose name corresponds to the deleted mount point.
◆
If the disk that is referenced by the mount point has been replaced, manually
re-create the mount point.
◆
If a file was not backed up due to a skip directive, you may still be able to select
the file for a recovery operation. In this case, although you can select the file, it is
not recoverable.
IMPORTANT
!
If you are recovering File Replication Services (FRS) or Windows NT Directory
Services (NTDS) data, boot the NetWorker client in Directory Service Recovery
mode before completing the recovery options in this section. “Booting
Windows in Directory Service Restore mode” on page 102 provides more
information about these recovery options.
System recovery in Windows Server 2008 and Windows Server 2003
When recovering an older backup of the System in Windows Server 2008 or Windows
Server 2003, any application or NMM upgrades, service packs, or hot fixes that were
installed since the backup may need to be reinstalled after the recovery. “Recovering
system data” on page 87 describes using the System Recover Session commands and
features for recovery.
Some applications have their own name for the recover session. When you select
Recover from the NMM client, it may display one or more of the following in
addition to System Recover Session, depending on which of the applications is
installed on the client:
◆
Active Directory Recover Session
◆
DPM Recover Session
◆
Exchange Recover Session
◆
Hyper-V Recover Session
◆
SharePoint Recover Session
Application-specific recovery steps are provided in separate chapters for each
application.
Recovering system data
The System Recover Session displays the backups that are available for recovery, and
provides browsing and searching to locate and select items for recovery.
To recover system data:
1. Open the NMM client software and select the NetWorker server on which the
NMM client software was configured for backup.
2. If the NMM client is part of a cluster, select the virtual client to which you are
recovering data. The virtual client can be selected from the Client list attribute in
the application toolbar.
3. From the left pane, select Recover > System Recover Session.
87
4. In the navigation tree, select the items to be recovered.
By default, items displayed in the navigation tree are from the most recent
backup. For the System Recover Session view, both snapshot and rollovers can be
displayed in the same view:
• If the most recent backup for an item is a snapshot, the snapshot is displayed.
• If the most recent backup for an item is a rollover, the rollover is displayed.
To recover file system items from a previous backup:
• From the application toolbar, click the Browse calendar icon and select an
earlier browse time.
• To view versions of a backup item prior to the browse time, right-click an item
in the navigation tree and select Versions from the pop-up menu.
5. To determine whether any volumes need to be mounted for a selected item,
right-click the item and select Required Volumes from the pop-up menu.
6. From the Options menu, select System Recover Session Options.
7. Enter application-specific recover options or settings, as described in:
• “NetWorker recovery options” on page 89
• “Exchange recovery options” on page 221
• “Performing DPM Server recovery” on page 258
• “Performing SharePoint 2007 recovery” on page 148
• “ADAM recovery options” on page 93
• “DFS recovery options” on page 94
• “FRS recovery options” on page 97
• “NTDS recovery options” on page 98
• “Performing SQL Server recovery” on page 121
8. From the System Recover Session toolbar, click Start Restore.
9. From the left pane, select Monitor to view messages related to the progress of the
recovery operation.
Note: If you recovered items from a snapshot (instant recovery), the snapshot volume will be
temporarily mounted during the recovery operation. The mounted volume is displayed in the
Windows Explorer program. After the recovery is complete, you may have to refresh the
Windows Explorer to update its view.
System Recovery Options Summary
This page lists the System Recover and NetWorker Recover Options. This allows you
to review the settings before starting the recovery.
To change the NetWorker and System Recover Options, click Recover Options.
You can also access the settings from the System Recover Session view. On the
System Recover Session toolbar click Recover Options.
The NetWorker Recover Options settings are specified on the General, NetWorker,
and Security tabs.
When the options are okay, click Start Recover to validate the options. If all option
settings are valid, the dialog box closes and recovery starts.
88
“Performing a recovery” on page 86 provides more information.
Recovery options
General recovery options:
• “General recovery options” on page 89
System recovery options:
• “NetWorker recovery options” on page 89
• “Cluster recovery options” on page 91
• “Exchange recovery options” on page 221
• “ADAM recovery options” on page 93
• “DFS recovery options” on page 94
• “Security recovery options” on page 100
General recovery options
This tab specifies how much diagnostic information is output for a recovery
operation. For most recoveries, the diagnostic information is displayed in the Monitor
view.
To set the level of diagnostic information for a recovery:
1. Click the General tab from the Recover Options dialog box.
2. From the Diagnostic Level attribute, select a level from 1 to 9.
A value of 1 produces the least amount of diagnostic information and a value of 9
produces the most diagnostic information. The value Normal produces no
diagnostic information.
3. Click OK.
Note: After each recovery, the Diagnostic Level attribute is automatically set to Normal.
NetWorker recovery options
The NetWorker tab is displayed in System Recover Session Options and Hyper-V
Recover Session Options. Some of the following features are not applicable to
Hyper-V Recover Session Options.
To specify recovery options for file system data:
2. Click the NetWorker tab.
3. In the Relocate recovered data to: field, type a location other than the location
from which the data was originally backed up. This feature is only applicable to
System Recover Session Options.
The location must be a local path. URL syntax such as \\servername\share is not
supported.
89
4. Use the Restore Type field to determine the default recovery method:
• Select Conventional Restore to recover from a conventional backup medium.
If a conventional backup is not available for the selected browse time, an
snapshot restore is performed.
• Select Snapshot Restore to recover from a snapshot. If a snapshot is not
available for the selected browse time, a conventional backup is performed.
For snapshot restore, snapshot validation will occur prior to the restore. If the
snapshot is invalid, the restore will fail. The invalid snapshot will be deleted
automatically the next time the snapshot group runs. You can also delete a
snapshot manually. “Delete a snapshot” on page 330 provides more
information about manually deleting snapshots.
5. Select Use Microsoft best practices for selecting the system state to enforce
Microsoft recommendations that determine which system state items must be
selected as a group for recovery. Clear this attribute to select system state items
individually. This feature is only applicable to System Recover Session Options.
!
CAUTION
Do not clear the "Microsoft Best Practices" box except where specifically
instructed to do otherwise in this document. For instance, there are some
procedures, such as recovering Distributed File System (DFS) data, where you
can clear this attribute.
Microsoft does not support recovery of portions of the operating system.
Attempted recovery of portions of the operating system, with this box cleared,
may render the system unbootable. After each recovery operation, this option
is automatically selected.
6. Select Terminate recover of item if errors are encountered if it is not already
selected. This checkbox is selected by default. If this option is selected, then when
an error occurs only the recovery of the specific item causing the error is halted.
For example, during the recovery of the items "C:\test" and
"APPLICATIONS:\SqlServerWriter," an error occurs during recovery of a single
file in the "C:\test" folder. If the checkbox is selected, the rest of the recovery of
"C:\test" will be terminated, but NMM will still attempt to recover
"APPLICATIONS:\SqlServerWriter\." If this box is cleared and there is a
recovery error with a file in "C:\test," NMM will still attempt to recover the rest
of "C:\test" and "APPLICATIONS:\SqlServerWriter."
Clear this checkbox if file system recovery fails, and you encounter the error
message Files failed to be restored for File system marked
objects.When this error occurs, open the PowerSnap client log file and note
which files caused the error. The location and name of this log file are described
in “PowerSnap client” on page 347.
7. Click OK.
“Windows system backup” on page 82 provides more information about recovering
data.
90
Cluster recovery options
Use this tab to specify whether to perform an authoritative or nonauthoritative
restore of the Cluster Writer.
If the Cluster Writer was backed up in NetWorker 7.4 SP2 prior to upgrading to
NMM 2.0, you must use a command line to perform an authoritative restore.
“Authoritative recovery from a NetWorker 7.4 SP2 save set after NMM upgrade” on
page 92 provides more information about the command line and when to use it.
Cluster Writer nonauthoritative restore (default)
Use this mode if the cluster is completely lost, in which case the cluster service does
not run on either node, and the cluster database is missing or corrupted.
Nonauthoritative restores will restore the cluster but will not restore a particular
version of the database. After the nonauthoritative restore is complete, you must
follow up with an authoritative restore to recover a specific version of the database.
NMM will perform an nonauthoritative restore by default, but you must make sure
that the SYSTEM COMPONENTS:\BootableSystemState save set is also selected.
To specify a nonauthoritative restore for the Windows Server 2008 Cluster Writer:
2. Click the NetWorker tab, and then clear the Use Microsoft best practices for
selecting system state checkbox.
3. Click OK to close the System Recover Session Options dialog box.
4. In the navigation tree, expand the SYSTEM COMPONENTS folder and select
Cluster Database.
5. Select the SYSTEM COMPONENTS:\BootableSystemState save set.
7. Reboot the system when prompted.
8. Repeat from step 1 on the other nodes, if needed.
Cluster Writer authoritative restore
Use this mode when the cluster configuration is lost, but the cluster is running okay
otherwise. For example, when a cluster resource was accidentally deleted or you
want to revert to a previous cluster configuration. The cluster must be healthy in all
nodes.
Note: Authoritative restores of the Cluster Writer are only supported for Windows Server 2008.
To specify an authoritative restore for the Windows Server 2008 Cluster Writer:
1. Make sure the cluster service is running on the local system.
2. Make sure that the cluster service is running on all remote nodes.
Microsoft Windows Server 2008 product documentation provides more
information on running the cluster service.
4. Click the NetWorker tab, and clear the Use Microsoft best practices for selecting
system state checkbox.
5. Click the Cluster tab.
91
6. Select the Authoritative Restore options.
Cluster Database.
Note: Perform an authoritative restore of the Windows 2008 cluster writer by selecting only the
Windows 2008 cluster writer in that recover session. In particular, do not perform an
authoritative restore of the Windows 2008 cluster writer at the same time as the restore of any
other NMM application.
Authoritative restores of the Windows 2008 cluster writer will restart the cluster service on all
nodes in the cluster.
Authoritative recovery from a NetWorker 7.4 SP2 save set after NMM upgrade
If the Windows Server 2008 cluster database was backed up in NetWorker 7.4 SP2
prior to upgrading to NMM 2.0, you must use a command line to perform an
authoritative restore. You cannot perform this authoritative recovery from the Cluster
tab of System Recover Session Options.
The following set of steps illustrates this installation, backup, and recovery sequence:
1. Install NetWorker 7.4 SP2 on a Windows Server 2008 cluster.
2. Back up VSS SYSTEM SERVICES.
3. Uninstall NetWorker 7.4 SP2.
4. Install NMM 2.0.
If you attempt to use the NMM recovery user interface to recover a cluster
database that was backed up in NetWorker 7.4 SP2 prior to upgrading to NMM
2.0, NMM will change it to a nonauthoritative restore.
You can only perform an authoritative restore of a Windows Server 2008 Cluster
Writer backup created with NetWorker 7.4 SP2 by using the command line.
To perform an authoritative restore of the Windows Server 2008 Cluster Writer:
1. Make sure that the cluster service is running in the local system.
2. Make sure that the cluster service is running on all remote nodes.
3. Type the following string on the command line:
recover -s NWServer -U -N "VSS SYSTEM SERVICES:\Cluster
Database"
Note: Perform an authoritative restore of the Windows 2008 cluster writer by selecting only the
Windows 2008 cluster writer in that recover session. In particular, do not perform an
authoritative restore of the Windows 2008 cluster writer at the same time as the restore of any
other NMM application.
Authoritative restores of the Windows 2008 cluster writer will restart the cluster service on all
nodes in the cluster.
Nonauthoritative recovery of a Windows Server 2008 cluster database with a NetWorker 7.4 SP2 save set
A nonauthoritative recovery of a Windows 2008 cluster database backup that uses a
save set created with NetWorker 7.4 SP2 can be performed through the NMM
recovery user interface or through the recover command line.
92
To specify a nonauthoritative recovery through the user interface by using a save set
created with NetWorker 7.4 SP2:
2. Click the NetWorker tab, and clear the Use Microsoft best practices for selecting
system state checkbox.
4. In the navigation tree, expand the VSS SYSTEM SERVICES folder and select
Cluster Database.
5. Select the entire VSS SYSTEM BOOT save set.
Repeat steps 1 through 7 for each node on the cluster as needed.
To specify a nonauthoritative recovery through the command line by using a save set
created with NetWorker 7.4 SP2:
1. Type the following command:
recover -s NWServer -N "VSS SYSTEM SERVICES:\Cluster
Database"
2. Type the following command:
recover -s NWServer -N "VSS SYSTEM BOOT:\"
Repeat steps 1 through 3 for each node on the cluster as needed.
ADAM recovery
options
Use the ADAM tab in the System Recover Session Options dialog box to specify
whether to perform an authoritative or nonauthoritative restore of Active Directory
Application Mode (ADAM) data on this host.
Note: In Windows Server 2008, Active Directory Lightweight Directory Services (AD LDS)
includes the ADAM functionality provided in Windows Server 2003. Unless otherwise noted,
assume that all references to ADAM in this chapter also apply to AD LDS.
ADAM nonauthoritative restore (default)
This type of restore is typically performed when the host is being recovered due to
catastrophic loss of data such as a disk failure. The most recent ADAM data is
restored from backups and then the recovered host is updated with the most recent
ADAM data from another ADAM server in the domain. This restore relies on at least
one other ADAM server in the domain to have accurate, up-to-date ADAM data.
ADAM authoritative restore
Perform an authoritative restore only when corrupted or deleted ADAM data has
been propagated to other ADAM server. This type of recovery restores the most
recent ADAM data from backups, and then updates other ADAM servers in the
domain with the recovered data.
Note: To recover ADAM data, the APPLICATIONS:\ADAM <instance_name> Writer\ save
set must have been backed up. Chapter 2, “Configuring a Scheduled Backup,” provides more
information about specifying save sets when configuring a backup.
93
To specify recovery options for ADAM data:
1. From the Options menu, select System Recover Session Options and then click
the ADAM tab.
2. Select one of the following options:
• Non Authoritative Restore
• Authoritative Restore
4. In the navigation tree, expand the Applications folder and select ADAM
<instance_name>Writer.
6. If an authoritative restore has been performed, restore for all selections has
completed, and NMM reports success, complete the following steps:
a. Exit the NMM client, and open a command window.
b. Run the dsdbutil.exe utility, and select Authoritative Restore option.
c. This marks the objects that are to be restored in authoritative mode.
d. When processing is complete, exit the dsdbutil.exe utility and start the ADAM
instance service manually.
If a nonauthoritative restore was performed, the ADAM instance has already
been started.
DFS recovery options
You can specify whether to perform an authoritative or nonauthoritative recovery of
Distributed File System (DFS) data on this host.
There are some differences between DFS recovery options on Windows Server 2003
and Windows Server 2008:
94
◆
Authoritative restore is available on Windows Server 2003 and on some Windows
Server 2008 systems.
◆
In Windows Server 2003, the writer is recovered from Applications:\DFS
Replication service writer.
◆
In Windows Server 2008, the writer is recovered from SYSTEM
COMPONENTS:\Bootable System State.
◆
In Windows Server 2008, DFS can have DFS shares or it can also be part of Active
Directory recovery.
◆
In Windows Server 2008, DFS is part of Active Directory recovery if it was
promoted to a Domain Controller with the “2008 Functional level” setting
selected.
◆
To restore DFS on a Windows 2008 system as part of Active Directory, the system
must be booted into Directory Services Restore Mode.
The steps for performing a DFS recovery from Windows Server 2003 and Windows
Server 2008 are described in separate procedures.
Nonauthoritative
restore (default)
This type of recovery restores the most recent DFS data from backups and then
updates the recovered host with the most recent DFS data from another DFS server in
the domain. A nonauthoritative recovery relies on at least one other DFS server in the
domain to have accurate up-to-date DFS data. Nonauthoritative recoveries are
typically performed when the host is being recovered due to catastrophic loss of data
such as a disk failure.
Authoritative restore
An authoritative recovery restores the most recent DFS data from backups, and then
updates other DFS servers in the domain with the recovered data. Perform an
authoritative recovery only when corrupted or deleted DFS data has been propagated
to other DFS servers.
Note: To recover DFS data, the APPLICATIONS:\DFS Replication service writer save set must
have been backed up. The DFS Replication service writer is available with Windows Server
2003 R2 or later. Additionally, the SYSTEM COMPONENTS:\ save set on the namespace server
(usually the domain controller) that is associated with the DFS configuration must have been
backed up. Chapter 2, “Configuring a Scheduled Backup,” provides more information about
specifying save sets when configuring a backup.
To specify which type of DFS recovery to perform in Windows Server 2003:
the DFS tab.
• Non Authoritative Restore to perform a nonauthoritative recovery.
• Authoritative Restore to perform an authoritative recovery.
4. From the navigation tree, expand the Applications folder and select DFS
Replication service writer.
6. After the DFS Replication service writer has been recovered, you must recover the
Bootable System State components on the namespace server (usually the domain
controller) that is associated with the DFS configuration:
a. Boot the namespace server in directory service restore mode as described in
“Performing a recovery” on page 86.
b. Open the NMM client software on the namespace server (usually the domain
controller) that is associated with the DFS configuration.
c. From the Options menu, select System Recover Configuration Options and
then click the NetWorker tab.
d. Clear the Use Microsoft best practices for selecting the system state attribute
and click Yes when asked to confirm your selection.
95
!
CAUTION
procedures, such as recovering Distributed File System (DFS) data, where
you can clear this attribute.
Attempted recovery of portions of the operating system, with this box
cleared, may render the system unbootable. After each recovery operation,
this option is automatically selected.
e. Click OK to close the dialog box.
f. From the left pane, select Recover > System Recover Session.
g. From the navigation tree, expand the SYSTEM COMPONENTS folder and
select BootableSystemState.
h. From the System Recover Session toolbar, click Start Restore.
After the recovery has completed, the Use Microsoft best practices for
selecting the system state attribute is automatically selected.
system data.
To specify which type of DFS recovery to perform in Windows Server 2008:
the DFS tab.
5. From the navigation tree, expand the SYSTEM COMPONENTS folder and select
BootableSystemState.
After the recovery has completed, the Use Microsoft best practices for selecting
the system state attribute is automatically selected.
96
!
CAUTION
FRS recovery options
File Replication Services (FRS) data on this host.
!
CAUTION
Boot the NetWorker client in Directory Service Recovery mode before completing
the recovery options in this section. Chapter 2, “Configuring a Scheduled Backup,”
provides more information about specifying save sets when configuring a backup.
FRS nonauthoritative restore (default)
A nonauthoritative recovery restores the most recent FRS data from backups and then
updates the recovered host with the most recent FRS data from another FRS server in
the domain. A nonauthoritative recovery relies on at least one other FRS server to
have accurate up-to-date FRS data. Nonauthoritative recoveries are typically
performed when the host is being recovered due to catastrophic loss of data such as a
disk failure.
FRS authoritative recovery
An authoritative recovery restores the most recent FRS data from backups, and then
updates the other FRS servers in the domain with the recovered data. Perform an
authoritative recovery only when corrupted or deleted FRS data has been propagated
to other FRS servers.
Note: To recover FRS data, the “SYSTEM COMPONENTS:\” save set must have been backed
up. Chapter 2, “Configuring a Scheduled Backup,” provides more information about
To specify whether to perform an authoritative or nonauthoritative recovery:
the FRS tab.
3. Click the NetWorker tab and then clear the attribute titled Use Microsoft best
practices for selecting the system state. Click Yes when asked to confirm your
selection.
97
!
CAUTION
5. From the navigation tree, expand the SYSTEM COMPONENTS folder and click
FRS.
If you are performing an authoritative recovery, complete the remaining steps in
“Performing an authoritative recovery of NTDS or FRS data” on page 103.
system data.
NTDS recovery options
Windows NT Directory Services (NTDS) data on this host.
!
CAUTION
Boot the NetWorker client in Directory Service Recovery mode before completing
the recovery options in this section. “Booting Windows in Directory Service
Restore mode” on page 102 provides more information about these recovery
options.
NTDS nonauthoritative restore (default)
A nonauthoritative recovery restores the most recent NTDS data from backups and
then updates the recovered host with the most recent NTDS data from another NTDS
server in the domain. A nonauthoritative recovery relies on at least one other NTDS
server to have accurate up-to-date NTDS data. Nonauthoritative recoveries are
typically performed when the host is being recovered due to catastrophic loss of data
such as a disk failure.
NTDS authoritative restore
An authoritative recovery restores the most recent NTDS data from backups, and
then updates other NTDS servers in the domain with the recovered data. Perform an
authoritative recovery only when corrupted or deleted NTDS data has been
propagated to other NTDS servers.
Note: To recover NTDS data, the “SYSTEM COMPONENTS:\” save set must have been
backed up. Chapter 2, “Configuring a Scheduled Backup,” provides more information about
98
To specify whether to perform an authoritative or nonauthoritative recovery:
the NTDS tab.
• Non Authoritative restore to perform a nonauthoritative recovery.
3. Click the NetWorker tab and then clear the attribute titled Use Microsoft best
practices for selecting the system state. Click Yes when asked to confirm your
selection.
!
CAUTION
procedures, such as recovering Distributed File System (DFS) data, where you
can clear this attribute.
Attempted recovery of portions of the operating system, with this box cleared,
may render the system unbootable. After each recovery operation, this option
is automatically selected.
5. From the navigation tree, expand the SYSTEM COMPONENTS folder and click
NTDS.
If you are performing an authoritative recovery, complete the remaining steps in
“Performing an authoritative recovery of NTDS or FRS data” on page 103.
system data.
99
Security recovery options
Backup and archive data on Windows hosts can be encrypted with an Advanced
Encryption Standard (AES) Application Specific Module (ASM) pass phrase. If no
pass phrase was specified when the backup data was encrypted, then the data is
encrypted with a default pass phrase. During data recovery, you must specify the
pass phrase used at the time of backup if it is not the default or current pass phrase.
To specify pass phrases:
the Security tab.
2. Type the pass phrases, and then click OK.
The EMC NetWorker Multiplatform Version Administration Guide, release 7.3.2 or later,
provides more information about AES encryption, and setting the pass phrase.
!
CAUTION
Do not use AES encryption when backing up files that are encrypted using
Windows Encrypting File System (EFS). When AES encryption is applied to a file
that is also encrypted using the Microsoft EFS, the backup will be reported as
successful. However, recovery of the file will fail.
100
Recovering file system snapshots that have not been rolled over
The procedure to recover items in a file system snapshot differs depending on
whether the snapshot has been rolled over to a conventional storage medium. This
section describes how to recover items from a file system snapshot that has not been
rolled over. “Windows system backup” on page 82 provides more information about
recovering file system snapshots that have been rolled over.
Note: If a snapshot volume has been mounted by using a client utility such as the EMC
NaviCLI interface or the Windows Disk Management utility, remove the mount point before
performing the next backup or recovery operation. If the mount point is not removed, it will be
removed at the end of the next NetWorker backup or recovery operation.
To recover items in a file system backup that has not been rolled over:
1. Connect to the NetWorker server on which the NetWorker client was configured
for backup.
“Opening the software and connecting to a NetWorker server” on page 43
provides more information about connecting to the NetWorker server.
3. In the navigation tree, right-click the snapshot and select Mount Snapshot.
If the Mount Snapshot option is unavailable, then the snapshot has been rolled
over. In this case, follow the steps described in “Windows system backup” on
page 82.
If another snapshot is currently mounted, a message indicates that the currently
mounted volume will be unmounted. Click OK to unmount the volume and to
proceed with mounting the selected snapshot volume.
4. When the taskbar animation stops and a green mark appears, click the mounted
snapshot.
The mounted snapshot item will expand in the navigation tree.
5. Under the expanded snapshot, select the items to be recovered.
selecting items for recovery.
6. Select any additional file system options if applicable.
“NetWorker recovery options” on page 89 provides more information about
these options.
Once the recovery starts, the snapshot is automatically unmounted.
8. From the left pane, select Monitor to view the progress of the recovery.
Recovering file system snapshots that have not been rolled over
101
Booting Windows in Directory Service Restore mode
Before you can complete the following recovery operations, boot the NMM client in
Directory Service Restore mode:
◆
DFS recovery
◆
FRS recovery
◆
NTDS recovery
◆
Active Directory (AD) disaster recovery
To boot the NMM client in Directory Service Restore mode:
1. On the NetWorker client domain controller, close all programs and restart
Windows.
When the computer restarts, a list of startup choices appears. The startup choices
and the duration of the startup display are based on the settings in the operating
systems section and the boot loader section of the boot.ini file.
2. Select the Windows boot option for the domain controller.
3. Press F8 to display a list of special boot options.
4. Select Directory Service Restore Mode (Windows Domain Controllers only)
from the list of special boot options.
When you boot in this mode, AD is taken offline.
5. Log in as administrator. Use the password that was specified when the domain
controller was created.
Windows starts in safe mode.
6. Open the NMM client application.
Continue with the recovery options. The following sections provide more specific
information:
102
◆
“DFS recovery options” on page 94
◆
“FRS recovery options” on page 97
◆
“NTDS recovery options” on page 98
◆
“Schema objects” on page 323
◆
“Windows system backup” on page 82 includes information about SYSTEM
STATE or VSS SYSTEM BOOT save sets, and save set recovery.
◆
“Performing a granular Active Directory or ADAM backup” on page 310 includes
information about save sets on a domain controller.
Performing an authoritative recovery of NTDS or FRS data
Perform an authoritative recovery only when corrupted or deleted data has been
propagated to other Windows NT Directory Services (NTDS) or File Replication
Services (FRS) servers. After the authoritative recovery, the domain administrator can
delete any unnecessary newer objects.
NTDS and FRS objects may have associated group policies, for example,
organizational units, domains, and site objects. Group policies are stored in the
SYSVOL directory. A recovery of the SYSVOL directory cannot be separated from an
authoritative recovery of NTDS or FRS.
To perform an authoritative recovery:
1. If you have not already done so, boot the NMM client in Directory Service
Restore mode.
“Performing a recovery” on page 86 provides more information about
performing this procedure.
2. If you have not already done so, complete the NTDS or FRS recovery options. The
following sections provide more specific information on the recovery options:
3. From the command line, use the xcopy command to copy sysvol\domain to
another location. This preserves group policy, and file and directory permissions.
4. Run the Windows ntdsutil utility.
5. At the ntdsutil prompt, type:
NTDSUTIL: authoritative restore
6. To perform an authoritative recovery of the following:
• Entire database, type:
NTDSUTIL: restore database
• Subtree or individual object, type:
NTDSUTIL: restore subtree distinguished_name
For example:
NTDSUTIL: restore subtree
OU=engineering,DC=Seattle,DC=jupiter,DC=com
NTDSUTIL: restore subtree
CN=mars,CN=users,DC=Seattle,DC=jupiter,DC=com
The Microsoft Windows Server Resource Kit provides more information and
Active Directory documentation.
7. Exit the ntdsutil utility by typing quit at each successive ntdsutil prompt until
the command prompt appears.
8. Copy the entire recovered SYSVOL\domain directory and all of its subdirectories
to a new location.
9. Restart the domain controller in normal mode.
10. Log in to the domain controller.
Performing an authoritative recovery of NTDS or FRS data
103
11. Wait for the SYSVOL share to be published.
This can take several minutes while the recovered domain controller
synchronizes with its replication partners.
12. After the SYSVOL share has been published, perform one of the following:
• For an authoritative recovery of the entire database, copy the entire
SYSVOL\domain directory tree from the new location to the existing
SYSVOL\domain directory as described in step 8 on page 103. When the copy
operation prompts for confirmation, select Yes to All.
• For an authoritative recovery of only a part of AD that includes Policy objects,
copy the policy folders from the new location (see step 8 on page 103) to the
existing policy folders. Policy objects can be identified by their global
universal ID (GUID). Policy folders are located in SYSVOL\domain\Policies.
104
Performing a directed recovery
In a normal recovery, there are three roles involved in the process: the source, control,
and destination roles. All of these roles are on the same computer:
◆
The source client is the computer where the backup data came from.
◆
The control client is the computer running the NMM UI to initiate the recover
process.
◆
The destination client is the computer that the data is restored to.
NMM supports normal recovery and directed recovery. In directed recovery, the source,
control, and destination roles are not all on the same computer. There are two types of
directed recovery:
◆
In pull directed recovery the control role and destination role are performed on the
same computer.The control role running on the destination computer pulls the
recovery data to itself. Backup data from the source client is restored to the
destination client.
◆
In push directed recovery the control role can be run on the source client or a
different client. Backup data from the source client is restored to the destination
client. The control client computer pushes it out to another computer, the
destination client. The destination client is not the same computer as the source
client or control client.
NMM supports pull directed recovery for most applications. You can run the
recovery program NMM (the control role) on the same computer that the data is
being restored to (the destination client). Some require additional or different steps.
Otherwise you can specify the alternate server to recover to in the regular recovery
steps for the application. NMM provides support for pull directed recovery for the
following applications:
◆
Microsoft Exchange Server 2007 “Microsoft Exchange Server 2007 directed
recovery” on page 238.
◆
Microsoft Office SharePoint Server 2007, within the same SharePoint farm. In
SharePoint, NMM provides directed recovery through recovery to an alternate
location. “Selecting alternate destination options” on page 183.
◆
File system “Performing a directed recovery” on page 105.
◆
Hyper-V “Hyper-V recovery to a different machine or location” on page 289. In
addition, NMM also supports push directed recovery for Hyper-V.
◆
Data Protection Manager (DPM), described in “Performing a granular recovery of
DPM objects” on page 260.
Peforming a directed recovery with NMM
To perform a directed recovery with NMM:
1. In NetWorker, create a volume backup. This will be the source for a directed
recover.
2. Run the NMM UI. Connect to the NetWorker Server that hosts the source NMM
client.
3. Add the NMM client that created the volume backup as a locally browsable
client.
105
Figure 10 on page 106 shows the default of only one NMM client available, the
local client.
Figure 10
Local client on the taskbar next to Client
4. On the Options menu, click Configure Options.
Figure 11 on page 106 shows the Configuration Options dialog box.
Refresh button
Figure 11
Configuration Options dialog box
5. Click the Refresh button, which is to the right of the Client name.
Figure 12 on page 107 shows the Select Viewable Clients dialog box. The clients
available on the NetWorker server that you are attached to are listed under
Available clients on.
106
Figure 12
Select Viewable Clients dialog box
6. Click the client to add in the Available clients on list box, and then click Add.
Add or remove clients as needed.
7. Click OK.
Figure 13 on page 107 shows the NMM window with the Client box.
Client list
Figure 13
NMM client dropdown list
8. Click the Client list, and select the client that created the volume backup as the
browsable client.
9. If the Snapshot Management or Monitor is the active navigation bar item, and
you are prompted with the Navigate Away dialog box, click Yes.
10. On the System Recover Sessions bar, select Recover Options.
The NetWorker System Recover Session Options dialog box appears.
11. On the NetWorker tab, specify the destination for the recovery in the Relocate
Recovered Data box, and then click OK.
12. Recover remote client's backup to local destination.
107
Recovering the Windows system configuration to an earlier state
Hosts that were upgraded from a regular NetWorker client to a NMM client can have
their Windows operating system configuration recovered to a state prior to
upgrading to the NMM client.
!
CAUTION
If the Windows operating system was upgraded on the host, you cannot recover the
Windows system configuration to the state it was in before the operating system
software was upgraded.
To recover the Windows system state to a point-in-time that precedes the installation
of the NMM client:
1. If the NMM client is a Windows domain controller, boot the NMM client in
Directory Services Restore mode.
“Booting Windows in Directory Service Restore mode” on page 102 provides
more information about booting in this mode.
If the NMM client is not a Windows domain controller, start with step 2 .
2. Recover the contents of the volume on which the Windows operating system was
installed.
3. Recover the Windows configuration save sets:
• If VSS was licensed and enabled, recover:
– VSS SYSTEM BOOT
– VSS SYSTEM FILESET
– VSS SYSTEM SERVICES
• If VSS was not licensed or enabled, recover:
– SYSTEM STATE
– SYSTEM FILES
– SYSTEM DB
4. Update the recovered registry with NMM client entries.
Because the registry was recovered to a client state prior to the NMM client
installation, you must update the registry with entries for the NMM client. To
update the registry, perform the following uninstall and install steps:
a. Uninstall the NMM client software.
b. Install the NMM client software.
c. Uninstall the NetWorker PowerSnap™ software.
PowerSnap is installed with the NMM client software.
d. Uninstall the NMM client software.
e. Reinstall the NMM client software.
Reinstall, as necessary, any other applications that were installed after the
point-in-time to which the Windows operating system configuration was
recovered.
The NetWorker Module for Microsoft Applications Release 2.2 Installation Guide
provides more information about the uninstall and install steps.
108
Disaster recovery
NMM provides recovery for the NMM client itself, as well as the applications NMM
protects, such as Microsoft Exchange Server, Microsoft SQL Server, Microsoft Office
SharePoint Services, Microsoft Hyper-V, and Microsoft Data Protection Manager.
A good disaster recovery requires a complete and timely backup. The procedures for
recovering an application include prerequisites of what particular files and data
should already be backed up when you need to perform the disaster recovery.
!
CAUTION
Ensure that your snapshot policies include at least one rollover to conventional
backup. Disaster recovery can only be performed from a conventional backup.
NMM requires conventional backup for SYSTEM COMPONENTS.
Attempting to restore SYSTEM COMPONENTS from a point-in-time backup can
leave the system in an unrecoverable state. NMM does not prevent an attempt to
restore SYSTEM COMPONENTS from a point-in-time backup. Chapter 14,
“Snapshot Management,” describes how to roll over a snapshot to a conventional
backup.
Disaster recovery of the entire NMM client machine requires completion of the
following tasks:
1. Recovery of the NMM client.
Perform one of the following, whichever is applicable:
• “Disaster recovery when the NMM client is not on a domain controller” on
page 110
• “Disaster recovery when the NMM client is on a domain controller” on
page 111
2. Full recovery of the applications on that machine.
Review these procedures when setting up backup plans, to ensure that the
prerequisite files and data will be backed up for future disaster recoveries.
Perform one of the following, whichever is applicable:
• “Performing SQL Server disaster recovery” on page 123
• “SharePoint 2007 disaster recovery” on page 170
• “Microsoft Exchange Server 2007 disaster recovery” on page 241
• “DPM disaster recovery” on page 267
• “Hyper-V parent partition disaster recovery” on page 292
• “Windows Server 2003 clustered client disaster recovery” on page 299
• “Cluster recovery options” on page 91
• “Active Directory disaster recovery” on page 324
Note: When performing a disaster recovery on a Domain Controller or Windows Server 2008,
the application must be reinstalled before performing disaster recovery of the NMM client.
Disaster recovery
109
Disaster recovery when the NMM client is not on a domain controller
To perform a disaster recovery if the NMM client is not on a domain controller:
1. Install the original operating system to the same specifications and configuration
as originally on the client.
Note: Be sure to apply all service packs and updates that had been applied to the operating
system before the disaster, to bring it up to the exact same level as existed at the time it was
lost.
2. Install the NMM client software exactly as it was configured before the disaster.
3. Open the NMM client software.
5. Select all local drives except for the NMM client installation folder and the
Replication Manager (RM) folder, typically C:\Program
Files\Legato\nsr\rmagentps.
6. Select the SYSTEM COMPONENTS folder.
If recovering a backup created with NetWorker VSS Client 1.0 SP1, under
SYSTEM COMPONENTS, clear ASR DISK.
8. When prompted, restart the client.
This restores the operating system to the state it was in when last backed up. Now
applications, such as Microsoft Exchange Server, SQL Server, Office SharePoint
Server, or Data Protection Manager, can be restored from previous backups.
9. To restore an application, follow the procedures for the specific application, listed
in “Disaster recovery” on page 109.
110
Disaster recovery when the NMM client is on a domain controller
To perform a disaster recovery if the NMM client resides on a domain controller:
1. Install the original operating system to the same specifications and configuration
as originally on the client.
Note: Be sure to apply all service packs and updates that had been applied to the operating
system before the disaster, to bring it up to the exact same level as existed at the time it was
lost.
2. Install the NMM client software exactly as it was configured before the disaster.
Note: This is for an Active Directory client NMM installation, and is not for the Active
Directory plugin within a client.
3. Boot the NMM client in directory service restore mode.
“Performing a recovery” on page 86 provides more information about these
recovery options.
4. Open the NMM client software.
6. Select all local drives except for the NMM client installation folder and the
Replication Manager (RM) folder, typically C:\Program
Files\Legato\nsr\rmagentps.
7. Select the SYSTEM COMPONENTS folder.
9. When prompted, restart the client.
This restores the operating system to the state it was in when last backed up. Now
applications, such as Microsoft Exchange Server, SQL Server, Office SharePoint
Server, or Data Protection Manager, can be restored from previous backups.
10. To restore an application, follow the procedures for the specific application, listed
in “Disaster recovery” on page 109.
Disaster recovery
111
112
5
Backup and Recovery
◆
◆
◆
◆
Microsoft SQL Server backup and recovery............................................................. 114
Performing SQL Server backups................................................................................ 119
Performing SQL Server recovery .............................................................................. 121
Performing SQL Server disaster recovery ............................................................... 123
113
Microsoft SQL Server backup and recovery
This chapter supplements the overall NetWorker Module for Microsoft Applications
(NMM) backup procedures in Chapter 2, “Configuring a Scheduled Backup,” with
the specific details needed to back up and recover Microsoft SQL Server.
Microsoft SQL Server supports Volume Shadow Copy Service (VSS) through
Microsoft-supplied application writers.
NMM allows the creation of a point-in-time copy of data. The snapshot includes exact
copies of data and open files. This allows the backup of the SQL Server while the SQL
database is online or in use.
SQL Server versions supported by NMM
NMM client supports the following versions of Microsoft SQL Server:
◆
Microsoft SQL Server 2008 (x86, x64):
• Standard
• Enterprise
◆
Microsoft SQL Server 2005 (x86, x64):
• Standard
• Enterprise
◆
Microsoft SQL Server 2008 Express edition
◆
Microsoft SQL Server 2005 Express edition
Note: NMM client does not support Microsoft SQL Server running on IA64.
The SQL VSS Writer does not support the encryption or compression features that SQL VDI
backup interface does. Therefore, NMM does not support these features.
SQL Server 2000 is no longer supported.
up-to-date lists of hardware, operating system, service pack, and application versions
supported by the NMM client.
Types of supported backup and recovery
Using point-in-time snapshot technology, NMM supports several types of backup
and recovery for Microsoft SQL Server. NMM supports database granularity for
snapshot backup and restore. This allows you to back up a particular SQL database
without having to shut down the database, and then select and restore that database
from the backed-up databases.
NMM provides the following types of backup:
114
◆
Only full level backups
◆
All SQL Server 2008 named instances and databases, including express and
stand-alone databases
◆
Only the SQL 2008 host instance and database name combination
◆
All SQL Server 2005 named instances and databases, including express and
◆
Only the SQL 2005 host instance and database name combination
NMM supports the following types of recovery:
◆
SQL Server 2008 stand-alone or express databases
◆
SQL Server 2005 stand-alone or express databases
Components used by NMM for SQL data backup and recovery
NMM uses the following writers and commands to back up SQL data:
◆
Application writers
NMM uses the application writers provided by SQL Server:
SqlServerWriter
The VSS writer for SQL Server 2005 and SQL Server 2008
Note: The MSDEWriter for SQL Server 2000 is no longer supported.
“Specifying save sets for SQL data” on page 115 describes how the writer name is
used in the save set syntax.
◆
Nsrsnap_vss_save
The NMM command to initiate the VSS-based backup.
◆
Nsrsnap_vss_recover
The NMM command to initiate the VSS-based recovery.
Specifying save sets for SQL data
Table 17 on page 115 lists the SQL save set syntax to specify for supported types of
SQL data. Specify SQL data save sets in the Save Set attribute of a Client resource.
Table 17
SQL save set syntax (page 1 of 2)
Type of data to back up
Save set syntax
All SQL Server 2008 named instances and
databases, including express and
If the SQL Server VSS Writer service is enabled (it is disabled by
default), type the following:
APPLICATIONS:\SqlServerWriter
Only the SQL 2008 host instance and
database name combination
APPLICATIONS:\SqlServerWriter\host%5Ci
nstance\master
For example, to back up the database named trans07 on the
SQL Server named instance MT11\BU, type the following:
APPLICATIONS:\SqlServerWriter\MT11%5CB
U\
trans07
Some special characters in a SQL Server named instance or
database name must be replaced with their URL-encoded
values. “URL encoding for SQL save sets” on page 117 provides
more information about URL encoding.
115
Table 17
SQL save set syntax (page 2 of 2)
Save set syntax
databases, including express and
database name combination
nstance\master
For example, to back up the database named trans07 on the
SQL Server named instance MT11\BU, type the following:
APPLICATIONS:\SqlServerWriter\MT11%5CB
U\
trans07
values. “URL encoding for SQL save sets” on page 117 provides
more information about URL encoding.
Displaying valid SQL data save sets
To display a list of the SQL save sets that are available for backup:
application data save set names:
• If the application server is not on a virtual host, type:
nsrsnap_vss_save -?
• If the application server is on a virtual host, type:
3. Press Enter.
The save sets that are available for backup are listed in a format similar to the
following:
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\master"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\model"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\msdb"
"APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\NADATA"
116
URL encoding for SQL save sets
◆
◆
two parts are always separated with a backslash (\) character. When specifying a SQL
Server named instance in the Save Set attribute, replace the backslash with its
URL-encoded value, which is %5C. In the following example, a SQL Server named
instance that is identified as:
CORPDBSERV1\YUKON
CORPDBSERV1%5CYUKON
If the database name in the previous example was NA#DATA, the save set entry
would be:
Table 18 on page 117 lists the most commonly used special characters and their URL
values.
Table 18
Special character
URL-encoded value
Special character
URL-encoded value
\
%5C
?
%3F
/
%2F
]
%5D
"
%22
[
%5B
%
%25
}
%7D
#
%23
{
%7B
&
%26
^
%5E
<
%3C
‘
%60
>
%3E
|
%7C
117
SQL Server application information variable settings
Information attribute of the Client resource.
Table 19
SQL variable settings for Application Information attribute
Attribute name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
This value is required.
application information attribute client that moves snapshots from
the NMM client to primary and
• The local host
client.
If setting up an SQL Client resource, it is
recommended that you use a proxy
client. If a proxy client is specified, SQL
consistency checks are performed on the
proxy client. Related messages are
written to the Replication Manager log
files (erm_clientxx.log) on the proxy
client.
If no proxy client is specified for an SQL
Client resource, consistency checks are
performed on the Client resource host.
Related messages are logged on the
Client resource host in both the
Replication Manager log files and the
nmm.raw file.
Serverless backup
serverless backups. Serverless backups free the application server from much of the
processing involved in a nonpersistent or instant snapshot backup. In a serverless
backup, a snapshot is created on the application server and then immediately moved
to the proxy client for further processing.
A serverless backup can be rolled over to a conventional backup, which can remain
on the mounted storage disk volume as a persistent snapshot.
118
Performing SQL Server backups
NMM supports several types and levels of backup:
◆
Point-in-time snapshots, which include exact copies of files and open files
◆
Full backup of SQL databases and logs
Note: EMC recommends that no more than 200 databases are backed up at one time. If
database sizes are large, this should be less.
Full backup of an SQL Server
The first four tasks for scheduling a backup are the same for all VSS writers that are
tasks in detail:
◆
◆
◆
◆
Task 5: Configure a SQL Server Client resource
The next task is specific to configuring a SQL Server Client Resource:
Configuration.
8. Select the Scheduled Backups attribute:
If needed, use URL-encoded values when specifying save set names in the Save
Set attribute of the Client resource.
URL-encoded values are used to represent special characters such as the
backslash (\). A table of values is provided in “URL encoding for SQL save sets”
on page 117.
To back up SQL Server 2005 and SQL Server 2008 named instances and
stand-alone databases, specify the save set name:
Performing SQL Server backups
119
The SQL 2005 and 2008 Express databases are backed up under SYSTEM
COMPONENTS:
SYSTEM COMPONENTS:\
The SQL Server save set syntax is described in “Specifying save sets for SQL
data” on page 115.
9. Specify SQL application variable settings in the Application information
attribute of the Client resource. These settings are described in Table 19 on
page 118.
11. In the Aliases attribute, type the NETBIOS name for the client.
Note: NMM client uses the host machine NETBIOS or “short” name when connecting to
the NetWorker server to browse backups. If the NETBIOS name is not found, NMM will
not be able to display backups.
120
Performing SQL Server recovery
Make sure that the SQL Server VSS Writer service was enabled during backup. SQL
Server 2005 and 2008 Express databases must be recovered from the SYSTEM
COMPONENTS:\ save set.
The SQL Server VSS Writer service is disabled by default when SQL Server is
installed.
Note: Though this release no longer supports SQL Server 2000, NMM can recover previous
backups of SQL Server 2000 data.
Recovering SQL Server 2005 and 2008 stand-alone databases from the SQLServerWriter save set
To recover SQL Server 2005 and 2008 stand-alone databases:
1. If restoring the master database, the SQL Server services for the SQL database
instances are automatically detected and stopped.
2. From the navigation tree, expand the Applications folder and the
SQLServerWriter folder.
3. Select the databases to recover.
5. After the recovery is complete, NMM automatically starts the SQL Server services
for the SQL Server database instances.
Recovering SQL Server 2005 or 2008 express databases from the System Components save set
To recover SQL Server express databases that were backed up to the SYSTEM
COMPONENTS save set:
1. If restoring the master database, stop the SQL Server services for the SQL Server
database instances.
2. From the navigation tree, expand the SYSTEM COMPONENTS:\ folder.
5. After the recovery is complete, start the SQL Server services for the SQL database
instances.
Best practices for SQL Server backup and recovery
Table 20 on page 122 lists the best practices and recommendations to follow when
using NMM to back up and recover SQL Server.
Performing SQL Server recovery
121
Table 20
122
SQL Server backup and recovery best practices and considerations
Consideration
Best practice
For application servers such as SQL servers, consider backing up the
backup group.
application server
program
mirroring
Split-mirror technology is supported only with the EMC VSS Provider.
the NMM client
If you attempt to use both the NMM client and a NetWorker module such
as NetWorker Module for Microsoft SQL Server to back up application
data, the module backups will be promoted to full backups. “NMM client
issues” on page 340 provides more information about client issues.
CLARiiON limits for the
CX500
If the number of snapshots exceeds the limit for number of snapshots
supported per CLARiiON, a "VETO" failure will occur during snapshot
creation. The CLARiiON limits for the CX500 are as follows:
For a CX500, the limit is 150 snapshots or snapshot sessions
system-wide. This includes reserved snapshots/sessions for SAN Copy.
There is also a limit of 8 snapshots or snapshot sessions per LUN.
Limits for CLARiiON systems can be found in the EMC SnapView
Integration Module for SQL Server (SIMS) Administrator’s Guide.
SQL instances in suspect
mode
If any of the databases belonging to a SQL instance are in suspect mode,
and a backup of the SQL instance is attempted with the Microsoft MSDE
Writer, the snapshot backup will fail.
Verify VSS SQL Writer
service is running
After installing SQL Server, verify that the VSS SQL Writer service is
running in order for the VSS SQL Writer snapshot to be successful. The
SQL Writer is disabled by default. It must be enabled or set to automatic
after installation.
Ignore tempdb database
Microsoft indicates that tempdb is a database that is rebuilt when the
server is rebooted, and that VSS Requestors can ignore this database.
The database is not reported by the writer and will not be backed up.
Special character
databases
Special character databases are not backed up and the best practice is to
have the list of databases without special character tables and columns.
Performing SQL Server disaster recovery
In cases where the SQL Server databases must be completely rebuilt, the NMM can
facilitate a disaster recovery. The following are scenarios and procedures for recovery.
SQL Server (standalone) disaster recovery on Windows Server 2008
Prior to disaster recovery, use NMM to perform regular backups of volumes, system
components, and application save sets.
The following save sets should be backed up regularly:
1. Back up the following with the snapshot policy “all.”
C:\
SYSTEM COMPONENTS:\
The backup of C:\ automatically includes the following:
• Windows Boot Volume
• Windows System Volume
Note: If the system has been set up with a separate file system for the Windows
boot\system partition and separate file system for application installation, then along with
C:\ drive, back up the file system that includes the Windows Boot Volume, Windows
System Volume, and application software.
“Specifying save sets for SQL data” on page 115 provides backup details.
2. Back up application data for SQL Server with the following save set:
To perform a disaster recovery of SQL Server on a Windows Server 2008 machine:
1. Install SQL Server and latest service packs, including the SQL Server instances
that were running on the machine before the disaster.
Note: The instance names must be exactly the same as the previous ones.
2. Perform a recover of the system components and volumes.
3. Perform a recovery of the SQL Server data for all previous instances.
Note: In the NMM interface, clear the checkmarks for all of the system databases, or the
recovery will fail.
SQL Server (standalone) disaster recovery on Windows Server 2003
C:\
SYSTEM COMPONENTS:\
123
“Specifying save sets for SQL data” on page 115 provides information for
SYSTEM COMPONENTS backup.
To perform a SQL Server (standalone) disaster recovery on Windows Server 2003:
1. Set up a Windows 2003 new machine, to match the machine name, IP address,
and domain status of the old machine.
2. Recover SYSTEM COMPONENTS:\ and the local file system volumes.
3. Reboot the machine.
4. Log in to the machine as local administrator, unjoin the domain, and then rejoin
the domain.
5. Recover the APPLICATIONS:\SqlServerWriter save set.
6. Reboot, and then recovery will be complete.
SQL Server cluster disaster recovery on Windows Server 2008
The SQL Server cluster should be backed up regularly, the following will be required
when a disaster recovery occurs:
1. Back up the following on both cluster nodes. Use the snapshot policy “all” for
each node backup:
C:\
SYSTEM COMPONENTS:\
124
Note: You must specify a separate applications save set for each SQL cluster instance.
To perform a SQL Server cluster disaster recovery on Windows Server 2008:
1. Set up each new cluster machine to match the machine name, IP address, and
domain status of the corresponding old machine.
2. On each cluster machine, install SQL Server 2005 and all updates that were
installed on the old machines, and all instances that were installed on the old
machines.
3. Recover each node’s SYSTEM COMPONENTS:\ and local file system volumes.
4. Reboot each new machine.
5. On each machine, log in to the machine as local administrator, unjoin the domain,
and then rejoin the domain.
6. On the active node:
a. Stop SQL Instance Services
b. Restore APPLICATIONS:\SQLServerWriter.
c. Start SQL Instance Services.
SQL Server cluster disaster recovery on Windows Server 2003
The SQL Server cluster should be backed up regularly, the following will be required
when a disaster recovery occurs:
1. Back up the following on both cluster nodes. Use the snapshot policy “all” for
each node backup:
C:\
SYSTEM COMPONENTS:\
2. Back up Cluster Management.
3. Back up application data for Microsoft SQL Server with the following save set:
Note: You must specify a separate applications save set for each SQL cluster instance.
125
To perform a Microsoft SQL Server two node cluster disaster recovery on Windows
Server 2003:
1. Set up each new cluster machine to match the machine name, IP address, and
domain status of the corresponding old machine.
2. Recover each node’s SYSTEM COMPONENTS:\ and local file system volumes.
3. Reboot each new machine.
4. On each machine, log in to the machine as local administrator, unjoin the domain,
and then rejoin the domain.
5. On each machine, recover the APPLICATIONS:\SqlServerWriter save set.
6. Reboot each machine.
Cluster status recovery is complete.
126
6
Microsoft SharePoint
Server Backup and
Recovery
◆
◆
◆
◆
◆
◆
◆
◆
◆
◆
◆
◆
SharePoint Server versions supported by NMM client .........................................
Types of SharePoint Server 2007 backup and recovery .........................................
SharePoint Server 2007 backup and recovery .........................................................
About optimized backup ...........................................................................................
Performing SharePoint 2007 backups ......................................................................
Task 5: Configure a SharePoint 2007 Client resource .............................................
Performing SharePoint 2007 recovery......................................................................
SharePoint 2003 backup and recovery .....................................................................
Performing SharePoint 2003 backups ......................................................................
Performing a SharePoint 2003 recovery...................................................................
SharePoint 2003 scenario of backup and recovery process ...................................
SharePoint 2007 disaster recovery ............................................................................
128
129
133
140
141
143
148
152
157
159
163
170
127
SharePoint Server versions supported by NMM client
The backup and recovery procedures are different for Microsoft Office SharePoint
Server 2003 and Microsoft Office SharePoint Server 2007. The information in this
chapter supplements the overall NMM backup procedures in Chapter 2,
“Configuring a Scheduled Backup.”
Microsoft Office SharePoint Server 2007 and Windows SharePoint Services 3.0
Backup and disaster recovery use the SharePoint VSS writers and Microsoft SQL
Server VSS writers. “SharePoint Server 2007 backup and recovery” on page 133
provides examples and steps for Microsoft Office SharePoint Server 2007 backup and
recovery. This section also applies to Windows SharePoint Services 3.0.
NMM also provides granular backup and recovery of SharePoint 2007 items.
Chapter 7, “Microsoft SharePoint 2007 Granular Backup and Recovery” provides
steps for performing granular backup and recovery.
Microsoft SharePoint Portal Server 2003 and Windows SharePoint Services 2.0
Backup and disaster recovery use the Microsoft SQL Server VSS writers. “SharePoint
2003 backup and recovery” on page 152 provides examples and steps for Microsoft
SharePoint Portal Server 2003 backup and recovery. This section also applies to
Windows SharePoint Services 2.0.
128
Types of SharePoint Server 2007 backup and recovery
NMM client provides these backup features:
◆
Backup types:
• Backup for stand-alone and distributed SharePoint farms
• Utilizes Volume Shadow Copy framework to create snapshot of SharePoint
data
◆
Full backup level, to provide:
• Full backup of SharePoint content databases
• Full backup of Search/Content Indexes
◆
Backup granularity:
• Farm backup, to back up all SharePoint data in the farm, including:
– Content/configuration databases
– Search/index databases
• Web application backup, to provide granular backup of single web
application
• SharePoint object, to provide granular backup of:
– Farm
– Site collection
– Site
– Sub-site
– List
Chapter 7, “Microsoft SharePoint 2007 Granular Backup and Recovery,”
describes NMM support for granular backup of web applications and
SharePoint objects.
NMM client provides these recovery features:
◆
Recovery types:
• Conventional
• Restore from snapshot
• Content recovery and disaster recovery
• Directed recovery in the same SharePoint farm
• Recovery for stand-alone and distributed SharePoint farms
Note: Rollback is unsupported.
◆
Recovery level — Full, content/configuration database, search/content indexes
129
◆
Advanced search — The backup data can be searched by:
• Document name
• Document type
• Date Range
• creation date
• Modification date
• Author
• Size
◆
Recovery granularity:
• Item-level recovery — Extensive granular support for List Libraries,
Site-Collection Templates, Web Applications, Site Collections, and Sites.
Chapter 7, “Microsoft SharePoint 2007 Granular Backup and Recovery,”
provides a list of SharePoint items available in NMM for granular recovery.
• Farm recovery:
– Content databases
– Configuration database
– Search/index databases
• Content database recovery — Restore of SharePoint content databases
Note: Any major changes made to the SharePoint configuration or database structure requires
a full backup. For example, if a full farm backup is performed on Monday, and then a new
content database is added on Tuesday, a full backup needs to be performed to keep the
backups up-to-date. The save set list also needs to be updated to include the new SharePoint
objects.
NMM solutions for protecting a Microsoft SharePoint farm
NMM 2.2 provides three solutions to protect a SharePoint farm:
◆
By using optimized backup and recovery — This method uses the SharePoint
Object Model to backup SharePoint web applications or site-collections.
SharePoint optimized backups allow backups at the SharePoint farm,
web-application, or site-collection levels. For these type of backups, the lowest
object level for recovery is a site-collection. This method provides a good balance
between recovery granularity and performance.
“About optimized backup” on page 140 provides information about optimized
backup.
◆
By using full granular backup and recovery — This method uses the SharePoint
Content Management API’s to backup SharePoint web applications,
site-collections, or sites for granular recovery. Both full and incremental, or
differential backups are supported. This allows recovery of SharePoint
applications, site- collections, sites, sub-sites, lists, libraries, calendars, wikis,
items, and nearly any other type of SharePoint content.
Chapter 7, “Microsoft SharePoint 2007 Granular Backup and Recovery,” provides
more information about granular backup and recovery for Microsoft SharePoint
2007.
130
◆
By using the SharePoint VSS Writer for disaster recovery — This method
leverages the Windows Server VSS framework to take VSS snapshots of the
SharePoint configuration database, search databases or indexes, and content
databases. These VSS snapshots can be backed up on-host using the Windows
VSS System Provider or off-host using VSS Hardware Providers. There is no
performance impact on the SharePoint farm except for the backup to disk or tape
operation, and this can be eliminated by using a proxy node and VSS Hardware
Provider for off-host and LAN-free backup.
“SharePoint 2007 disaster recovery” on page 170 provides details about using
SharePoint VSS writer for disaster recovery.
Table 21 on page 131 provides a comparison of VSS, optimized, and granular
backups.
Table 21
Comparison of VSS, optimized, and granular backups (page 1 of 2)
SharePoint data protection
Feature
Support provided
VSS backup
Optimized
backup
Granular
backup
SharePoint application support
Applications
SharePoint Server 2007
X
X
X
WSS 3.0
X
X
X
SharePoint Server 2003
X
WSS 3.0 / WSS 2.0
X
Global source based
deduplication
X
X
X
Target based deduplication
X
X
X
Traditional LAN
X
X
X
Off-host via proxy
X
LAN-free (SAN)
X Proxy
Farm
X
Databases
X
Full
X
X
X
SharePoint disaster recovery
Deduplication
Topologies
Backup support
Backup types
Incremental (log)
X
Differential
Recovery
Farm
X
CFG Database
X
Search database/index
X
Content databases
X
131
Table 21
Comparison of VSS, optimized, and granular backups (page 2 of 2)
Optimized
backup
Granular
backup
Web applications
X
X
Site collection
X
X
SharePoint data protection
Feature
Support provided
SharePoint granular recovery
Recovery
VSS backup
Sites
X
Lists (calendars, events, surveys,
wiki's, blog's, announcements,
and so on.)
X
Libraries (document, picture, and
so on.)
X
Items
X
Versioning
X
Metadata (Author, Timestamp,
and so on)
X
Search
Author, Name, Title, Size,
Creation Date, Modification Date,
Keyword, Save set Date Range
X
Directed recovery
In-place recovery (same farm)
X
X
Out-of-place recovery (same
farm)
X
X
Farm to farm recovery
X
X
In this table, X denotes supported.
132
SharePoint Server 2007 backup and recovery
This section describes NMM support for backup and recovery of Microsoft Office
SharePoint 2007, and Windows SharePoint Services 3.0 that use SharePoint 2007
writers.
Storage requirements and performance impact in SharePoint backup
NMM performs the backup process in stages, so that you do not need as much free
space for the backup as the data in the SharePoint databases. During backup, the
objects are staged on the web front-end, and then sent to the NetWorker server. The
data staging operations and sending to the NetWorker server operations occur in
parallel.
Example of staging process
In this example, a SharePoint Farm has 8 top level sites, Site 1 to Site 8. Each site has a
maximum size of 0.25 TB, so the total database size is 2 TB. To backup this SharePoint
farm, only 0.25 TB space is needed on the web front-end:
1. Site 1 is staged to a local disk on the web front-end. This consumes a maximum of
0.25 TB of disk space.
2. Once the staging for Site 1 is completed, the staged data is sent to the NetWorker
server.
3. After sending the staged data to the NetWorker server, the consumed space is
reclaimed and is free disk space again.
4. Site 2 is staged, and steps 2 and 3 are performed again. This repeats for each site
until all are backed up to the NetWorker server.
Before staging of any site, the NetWorker SharePoint Service Controller service
calculates the available free disk space and required disk space for the staging. If the
available disk free space is less than the space required for staging, NetWorker
SharePoint Service Controller service waits for nsr_moss_save.exe to free the used
space. Once nsr_moss_save.exe frees the temporary staging after sending data to the
NetWorker Server, NetWorker SharePoint Service Controller service starts staging
the site.
SharePoint backup performance tips
◆
Use frequent incremental backups between full backups to reduce the need for
full backups. Incremental backups require less network, processor, and data
storage resources.
◆
To decrease wait time for the NetWorker SharePoint Service Controller service,
and increase system performance on the web front-end, provide as much free
disk space as you can.
◆
A public network is used to transfer data from the SQL Server back end to the
web front-end during data extraction. To improve data transfer rates and
network performance, install a web front-end and the NMM client on the system
where the SQL Server resides, and use this web front-end for backup and restore
purposes only.
133
NMM installation requirements for a Microsoft Office SharePoint 2007 system
A Microsoft Office SharePoint 2007 farm can be deployed in a stand-alone system.
This system runs all services on one machine as shown in Figure 14 on page 134, or in
a configuration that includes several servers that host separate services, as shown in
Figure 15 on page 135. For NMM to backup a SharePoint farm, the NMM client must
be installed on each machine in the farm.
A SharePoint farm includes the following services:
◆
SharePoint 2007 Servers:
• Central Administration site and shared services — This is usually installed on
an application server.
• Front-end web servers — This is the web page-based user interface to manage
the server.
• Index server — Can be included on the query server if there is only one query
server.
• Query servers — If there is more than one query server, the index server
cannot be included on a query server.
• Application servers — For example, Excel Calculation Services.
• Microsoft SQL Server — This server contains the SharePoint databases:
– Configuration database (only one per farm)
– Content databases (one or more per farm)
– Search database (one or more per farm)
!
IMPORTANT
To back up the entire farm, the NMM client must be installed on each server that
hosts SharePoint data, and at least one front-end server.
Configuration
Database
Content
Database
NetWorker Server
7.3.3 or later:
Administers protection
for NetWorker clients
such as NMM
Web Front End
Index Server
Query Server
GEN-000751
Figure 14
134
SharePoint 2007 stand-alone farm configuration
Query Server
Index Server
If there is more than Query
Server, the Query Servers
cannot be included on the
Index Server
Configuration Database
(1 per farm)
NetWorker Server
7.3.3 or later:
Administers
protection for
NetWorker clients
such as NMM
Content Database 1
(1 or more per farm)
Web Front End
On an application server
Content Database 2
Contains all SharePoint
databases
GEN-000752
Figure 15
SharePoint 2007 distributed farm configuration
SharePoint 2007 data backed up by NMM
NMM client backs up the following SharePoint components:
◆
Configuration database, which is a SQL configuration database
◆
Content databases, which are SQL content databases
◆
SharePoint Search, which includes SharePoint Search Index and associated SQL
database
◆
Office Search, which includes Office Search Index and associated SQL database
Note: The Search and Index databases cannot be backed up by themselves. Search/Index
databases are backed up and recovered as part of farm-level backup and recovery.
Keeping backups and recoveries in sync
Set up a consistent schedule of full farm-level backups. If backing up individual
content databases, schedule these in between the full farm-level backups. However,
whenever there are significant changes to the SharePoint farm, you should
immediately update the NetWorker backup configuration with any new SharePoint
objects, and perform a new full farm-level backup as soon as possible.
135
Configuration or database structure changes
Any major changes made to the SharePoint configuration or database structure
requires an update of the save set lists and a full backup. For example, if a full farm
backup is performed on Monday, and then the user adds a new content database on
Tuesday, a full backup needs to be performed to keep the backups up-to-date. For the
full backup to be up-to-date, the NetWorker backup configuration must be manually
updated to include the new database save sets.
Out-of-sync expiration policies between clients
In a distributed farm, a snapshot policy is specified at the group level, to apply
uniformly to all SharePoint clients in a group. The snapshot policy is then applied to
each client independently. As long as the farm is working properly in a steady state,
the snapshot expiration policy for the group and for each client stay in sync. But if
one of the clients fails, and you do a manual rollover or a delete on one of the hosts,
the rest of the hosts fall out-of-sync. When they get out-of-sync, some of the
snapshots will expire and a full recovery is no longer possible from those snapshots.
To bring the snapshots up-to-date, run a new full backup. A full backup of a
distributed farm requires special steps in configuring the Client resources and
running the backup:
◆
“Configuring a SharePoint Client resource to perform a full backup of a
distributed farm” on page 144
◆
“Performing a full backup for a distributed SharePoint farm” on page 148
Components used by NMM for SharePoint backup and recovery
NMM uses the following VSS writers provided by Microsoft for SharePoint 2007, as
well as a writer to back up the SQL database:
136
◆
SqlServer Writer, backs up the SQL databases.
◆
SharePoint Services Writer, the writer for SharePoint 2007, which runs on the
Web front end.
◆
SharePoint Search Writer, the writer for SharePoint 2007 search.
◆
Office Server Search Writer, the writer for Microsoft Office Server search.
Specifying SharePoint 2007 save sets for application data
Table 22 on page 137 lists the SharePoint 2007 save set syntax that specify the
supported types of SharePoint 2007 data. Specify the SharePoint 2007 data save sets
in the Save Set attribute of a Client resource.
Table 22
SharePoint save set syntax
Type of backup data
Save set syntax
SharePoint databases on
SQL Server
Full SharePoint backup
The SharePoint Search Index
and SharePoint Content Index
are included when this save
set is specified.
APPLICATIONS:\Microsoft Office SharePoint
Services
SharePoint content database
A specific content database.
For example, on the host
Mars1, a database named
generic_database_1
Services\Mars1\generic_database_1
SharePoint Search Index and
SharePoint Content Index
SYSTEM COMPONENTS:\Microsoft Office SharePoint
Search
or
SYSTEM COMPONENTS:\Microsoft Office Search
Note: The Search and Index databases cannot be backed up by themselves.
Search/Index databases are backed up as part of a Farm-level SharePoint
backup.
Note: The Search and Index save set names are provided for information only.
These databases cannot be backed up by themselves. Search/Index databases
are backed up as part of a farm-level SharePoint backup.
Displaying valid SharePoint 2007 data save sets
To display a list of the SharePoint 2007 save sets that are available for backup:
If the application server is not on a virtual host, type:
nsrsnap_vss_save -?
If the application server is on a virtual host, type:
node that currently hosts the application server.
137
3. Press Enter.
APPLICATIONS:\Microsoft Office SharePoint Services
Set attribute of a Client resource. Each entry that you add to the Save Set
attribute must be typed on a separate line.
URL encoding for SharePoint save sets
characters when naming SQL or SharePoint 2007 save sets. Table 23 on page 139 lists
the most commonly used special characters and their URL values.
Note: Save sets displayed by the nsrsnap_vss_save - ? command already include the URL
encoding.
The name of a SQL Server named instance or SharePoint Server content database
consists of two parts: the hostname and a unique name. These two parts are always
separated with a backslash (\) character. When specifying a SQL Server named
instance in the Save Set attribute, replace the backslash with its URL-encoded value,
which is %5C. In the following example, a SQL Server named instance that is
identified as:
CORPDBSERV1\YUKON
CORPDBSERV1%5CYUKON
If the database name in the previous example was NA#DATA, the “#” would be
replaced with the URL-encoded value “%23”. The save set entry would be:
When specifying a SharePoint Server content database in the Save Set attribute,
replace the backslash with its URL-encoded value, which is %5C. In the following
example, a SharePoint Server content database that is identified as:
Mars1\generic_database_1
Mars1%5generic_database_1
To back up only the database named “generic_database_1” on the server “Mars1” the
complete entry in the Save Set attribute would be:
Services\Mars1%5generic_database_1
138
Table 23
Special character
URL-encoded value
Special character
URL-encoded value
\
%5C
?
%3F
/
%2F
]
%5D
"
%22
[
%5B
%
%25
}
%7D
#
%23
{
%7B
&
%26
^
%5E
<
%3C
‘
%60
>
%3E
|
%7C
SharePoint 2007 application information variable settings
Table 24
SharePoint application information variable settings
Attribute name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
application information
attribute
Client that moves snapshots from
the NMM Client to primary and
• The local host
client.
139
About optimized backup
Optimized backup is taken when you want to restore a complete web application and
site collection. The performance of NMM 2.2 optimized backups is much better than
granular backups because the use of optimized backup drastically reduces the
backup window for large SharePoint site backups. Using optimized backup, you can
not do recovery at a granular or object level. Optimized backups can only be restored
in full. There can be no partial or item-level restores from an optimized backup.
Using optimized backups, a web application and site collection can be recovered back
to the original location or to an alternate location in the same SharePoint farm or to a
different farm. Optimized backup is always performed at full level because
incremental backup is not supported.
Optimized backup is not the default backup type that is supported. In order to
perform optimized backup the parameter “Optimized =True” should be set in the
Application information of the Client resource in NMC.
Perform optimized backup and recovery
For optimized backup, the backup can be scheduled using the NMC GUI, or started
manually from command line.
By using the NMC GUI
To perform backup and recovery by using NMC:
1. Create a client resource and specify the save set to back up in the Saveset
attribute. The save set can be obtained by running the nsr_moss_save command
on the NMM client web front-end where NMM is installed with the options
described in Table 25 on page 140.
Table 25
Save sets for backup
Command
Results
nsr_moss_save –f
This lists the name of the SharePoint farm. Two parts are listed. One is the URL for
the site, and the other is the save set that needs to be copied and pasted into the
Save Set field in NMC.
nsr_moss_save –w
This lists all the web applications available in the farm. Two parts are listed. One is
the URL for the web application, and the other is the save set that needs to be
copied and pasted into the Save Set field in NMC.
nsr_moss_save –w <Url for
Web application>
This lists all the top level sites available in that web application. Two parts are
listed. One is the URL for the web application, and the other is the save set that
needs to be copied and pasted into the Save Set field in NetWorker NMC.
2. After successful backup, start the NMM GUI in the web front-end. The GUI lists
all the backed up objects. The GUI does not display the objects in the granular
level when optimized backup is performed.
3. Select the object to be recovered and click the Recovery button. The recovery
location can be the original location or to an alternate location.
4. In the Backup Command tab, specify nsr_moss_save. In the Application
Information tab, specify Optimized=True.
5. Create a group and assign the client to the group created. In the Group resource,
choose the option to start the backup at a convenient time.
140
Command line backup or manual backup
The command line backup can be performed in web front-end. To perform a full
optimized backup, run the following command:
nsr_moss_save -s <server name> -c <client name> -N <saveset> -l
full -A Optimized=True
Performing SharePoint 2007 backups
NMM backup of SharePoint provides:
◆
Point-in-time snapshots, which include exact copies of files and open files.
◆
Full SharePoint Server backup.
◆
SharePoint content and configuration database backup.
Configuring SharePoint 2007 backups
The NMM client must be installed on each machine in the SharePoint farm. In
addition, for distributed configurations, the SharePoint writer must be registered on
on one, and only one, of the web front-end servers. This web front-end server will be
included in the NetWorker backup schedule, and will be used to perform any
recoveries.
All of the following services must be enabled and started on the computers where the
VSS writers will be run, or else backup will fail:
◆
SharePoint Services Writer, on the web front-end computer only
◆
SQLServer Services VSS Writer, on any computer that contains the Configuration
database or one or more content databases
In addition, the following services should already be enabled and started on the
computers where SharePoint search activities are being performed. There will be VSS
Writers associated with them:
◆
Office SharePoint Server Search
◆
Windows SharePoint Server Search
Note: On a stand-alone farm, all of these writers and services run on one computer. In a
distributed farm, the SQL Server and SharePoint Servers may run on separate computers. The
service for the writers must be enabled on the computer running the service.
To register SharePoint writer, type the following at the command line:
STSADM.EXE –o registerwsswriter
The default location for STSADM.EXE is in the following folder:
C:\Program Files\Common Files\Microsoft Shared\web server
extensions\12\BIN
To start the SharePoint and SQL Server writers, complete the following steps on the
computer where the writer is registered. To determine the writers available on a
specific machine, follow the steps in “Displaying valid SharePoint 2007 data save
sets” on page 137.
141
Complete the following steps on each machine where the VSS writers will be run:
1. Start Windows Services by using the appropriate steps for your operating system:
a. In Windows Server 2003, click Start, click Control Panel, click Administrative
Tools, and then double-click Services.
b. In Windows Server 2008, click Start, point to Administrative Tools, and then
click Services.
2. Right-click one of the following services, and then click Properties:
• SharePoint Services Writer (on the web front-end machine only)
• SQLServer Services VSS Writer (on any machine that contains the
configuration database or one or more content databases)
3. In Startup type, click Automatic, and then click OK.
4. Right-click the writer you just enabled, and then click Start.
5. Repeat step 2 through step 4 for each of the listed writers present on the
computer.
6. Repeat step 1 through step 5 as needed on each computer in the farm to enable
and start the appropriate writers on each computer.
Once these writers are registered and started, you can schedule a backup. The first
four tasks for scheduling a backup are the same for all VSS writers supported by
NMM. Chapter 2, “Configuring a Scheduled Backup,” describes these tasks in detail:
◆
◆
◆
◆
NMM provides full backup, and backup of individual content databases, which
allows you to protect your SharePoint farm with individual content database
backups between full backups. Consider configuring separate backup schedules for
full backups and individual content database backups.
The following task provides specific steps for SharePoint 2007:
◆
142
“Task 5: Configure a SharePoint 2007 Client resource” on page 143
Task 5: Configure a SharePoint 2007 Client resource
For each machine in the SharePoint farm, create a Client resource in NetWorker
Management Console, and assign it to the same backup group. The steps to create the
Client resource depend on whether it is a stand-alone SharePoint farm, distributed
SharePoint farm, and a full or component backup:
◆
“Configuring a SharePoint Client resource to back up a stand-alone farm” on
page 143.
◆
“Configuring a SharePoint Client resource to perform a full backup of a
distributed farm” on page 144.
◆
“Configuring a SharePoint Client resource to back up a component on a machine
in a distributed farm” on page 146.
There are also special steps when running the distributed SharePoint farm backup, as
described in “Performing a full backup for a distributed SharePoint farm” on
page 148.
Configuring a SharePoint Client resource to back up a stand-alone farm
In a stand-alone SharePoint configuration, only one component is required in the
save set to perform a full backup.
To configure a SharePoint Client resource to back up a stand-alone farm:
Configuration.
entries on separate lines.
To back up all SharePoint Server data, where all server components are located
on the same machine, specify the save set name:
The SharePoint Server save set syntax is described in “Specifying SharePoint 2007
save sets for application data” on page 137.
143
10. If needed, use URL-encoded values when specifying save set names in the Save
backslash (\). A table of values is provided in “URL encoding for SharePoint save
Note: If a save set is copied from the nsrsnap_vss_save - ? output, it will already include
the URL encoding.
will be added.
If Client resources for the same NMM Client host are added to different backup
NSR_SNAP_TYPE=vss
Type SharePoint application variable settings and their values on separate lines.
These settings are described in Table 24 on page 139.
Note: NMM Client uses the host machine NETBIOS or “short” name when connecting to
the NetWorker Server to browse backups. If the NETBIOS name is not found, NMM will
19. Click OK.
Configuring a SharePoint Client resource to perform a full backup of a distributed farm
In a distributed configuration, you must set up a Client resource for the web
front-end server where the SharePoint 2007 VSS Writer was enabled using
STSADM.exe. You do not need a Client resource on each web front-end unless you
are implementing a system-level disaster recover backup schedule for these hosts.
The Client resource can specify to perform a full farm backup, or to back up a content
database. The components to specify in the save set for a particular machine are
determined by performing the steps in “Displaying valid SharePoint 2007 data save
sets” on page 137 on that machine.
To configure a SharePoint Client resource to perform a full backup of a distributed
farm:
1. Repeat step 2 on page 145 through step 20 on page 146 for each machine in the
farm, specifying the appropriate save set for the machine in step 10 on page 145 .
For a distributed farm backup, add all client resources for the farm members to
the same NetWorker backup group.
144
Configuration.
retention policy determines how long rolled-over data is available.
10. In the Save Set attribute, specify the components to be backed up:
Note: Each machine in a distributed farm requires a separate Client resource, and the
Client resource for a given machine should only contain the SharePoint save sets on that
machine.
• For the machine where the SharePoint VSS writer is registered, specify the
save set name:
Then proceed to step 11 .
• For the SQL Server machine, specify the save set name:
• For all other SharePoint machines, determine the save sets available on the
specific machine by following the steps in “Displaying valid SharePoint 2007
data save sets” on page 137. Specify the available save set names for that
machine.
Search/Index databases are backed up as part of a Farm-level SharePoint backup.
Note: If a save set is copied from the nsrsnap_vss_save - ? output, it will already include
the URL encoding.
145
will be added.
For a distributed farm backup, add all client resources for the farm members to
the same NetWorker backup group.
16. In the Backup command attribute, type the backup command.
For all types of backup data except for Active Directory conventional backups,
type:
NSR_SNAP_TYPE=vss
20. Click OK.
After the Client resources have been configured on each machine, the backup process
itself requires special steps as described in “Performing a full backup for a distributed
SharePoint farm” on page 148.
Configuring a SharePoint Client resource to back up a component on a machine in a distributed farm
To configure a SharePoint Client resource to back up a component on a machine in a
distributed farm:
Configuration.
146
• To back up all SharePoint databases (Configuration database and Content
databases), specify the save set name:
Note: The Configuration database cannot be backed up by itself.
Then proceed to step 10 on page 147 .
• To back up a specific Content database, specify the save set name:
APPLICATIONS:\SqlServerWriter\<database name>
Then proceed to step 10 on page 147 .
Search/Index databases are backed up as part of a farm-level SharePoint backup.
will be added.
type:
NSR_SNAP_TYPE=vss
These settings are described in “SharePoint application information variable
settings” on page 139.
147
19. Click OK.
Performing a full backup for a distributed SharePoint farm
In a distributed SharePoint farm, one or more SharePoint services may be located
remotely on separate machines from the SharePoint web front-end. In order for
NetWorker to successfully save all Client resources in the SharePoint farm, backups
must be initiated through the Client resource on each remote machine, as well as the
web front-end, and all Client resources must be in the same NetWorker group. If the
backup is only initiated through the web front-end, recovery will fail because even
though NMM displays all of the SharePoint farm save sets on the web front-end, the
remote client backups were never performed.
To perform a full backup for a distributed SharePoint farm:
1. Configure a Client resource for each machine, as described in “Configuring a
SharePoint Client resource to perform a full backup of a distributed farm” on
page 144.
Note: All clients in the farm must be part of the same backup group.
2. Run the NetWorker group to back up the data for all of the members of the farm.
To perform a recovery after a full backup, it must be initiated from the web front-end.
In that process, NMM prompts you to recover each remote machine before
recovering the web front-end machine. These steps are described in “Full recovery of
a distributed SharePoint farm” on page 150.
Performing SharePoint 2007 recovery
After the backup is complete, NMM truncates the files. When the files are backed up,
administrators can perform:
148
◆
“Full recovery of a stand-alone SharePoint farm” on page 149
◆
“Full recovery of a distributed SharePoint farm” on page 150
◆
“Recovery of individual SharePoint content databases” on page 151
◆
“Recovery of SharePoint Search/Content Indexes” on page 151
◆
“SharePoint 2007 disaster recovery” on page 170
◆
Chapter 7, “Microsoft SharePoint 2007 Granular Backup and Recovery”
Mandatory order for SharePoint recoveries
In all SharePoint recoveries, the recovery process must be performed in the following
sequence:
1. Configuration database
2. Content databases
Note: Data loss will occur if this sequence is not followed in order.
Configuration database recovery requirements in full farm recovery
When a SharePoint Configuration database is recovered as part of the recovery of an
entire farm, all Content databases in that farm must also be recovered in order for the
Microsoft SharePoint Writer to ensure consistency. Previously this was considered a
best practice, but Microsoft states that this is a requirement.
An individual Content database can still be recovered if the recovery is not part of an
entire farm recovery. “Recovery of individual SharePoint content databases” on
page 151 provides steps for recovering individual SharePoint Content databases.
Selecting valid items for recovery
All valid SharePoint items available for recovery are displayed in the
APPLICATIONS folder in NMM. A SYSTEM COMPONENTS node may be listed in
NMM after a SharePoint backup, but it might not contain valid savesets. If SharePoint
items appear under both the APPLICATIONS folder and SYSTEM COMPONENTS
node, select the recovery items from the APPLICATIONS folder.
Full recovery of a stand-alone SharePoint farm
To perform a full recovery of Microsoft Office SharePoint Portal Server 2007:
1. Open NetWorker Module for Microsoft Applications.
2. In the navigation tree, expand the APPLICATIONS folder and select SharePoint
Services Writer.
Recovery proceeds to completion. Details about the recovery appear in the
monitoring window.
149
Full recovery of a distributed SharePoint farm
A full recovery of a distributed SharePoint farm requires that each machine in the
farm is configured as a Client resource, and each remote machine was backed up
before the web front-end machine was backed up. These steps are described in
“Performing a full backup for a distributed SharePoint farm” on page 148.
To perform a full recovery of Microsoft Office SharePoint Portal Server 2007:
1. Open NetWorker Module for Microsoft Applications on the web front-end that
was used to perform the backups of the SharePoint 2007 Writer.
2. In the navigation tree, expand the APPLICATIONS folder and select SharePoint
Services Writer.
The Restore Remote Dependencies dialog box appears, alerting you to go to the
specified hosts and start the recovery processes.
Figure 16 on page 150 shows an example of several remote hosts that must be
recovered. In this example, though 13 service components are displayed, they are
located on three remote hosts. You would need to go to each of those three hosts
once to recover all 13 services.
Figure 16
Restore Remote Dependencies dialog box
4. At each remote host listed in the Restore Remote Dependencies dialog box:
a. Open NetWorker Module for Microsoft Applications.
b. In the navigation tree, locate and mark the items.
c. Restore the items.
5. After all remote hosts have been recovered, return to NetWorker Module for
Microsoft Applications on the web front-end and click Continue in the Restore
Remote Dependencies dialog box. Recovery proceeds to completion. Details
about the recovery are displayed in the monitoring window.
150
Recovery of individual SharePoint content databases
One or more content databases can be recovered after the configuration database has
been restored.
In the previous releases of NMM, a user was unable to select only individual
SharePoint content databases for restore. When any content database was selected,
the corresponding configuration and generic databases were also selected for
recovery. This release of NMM provides the ability to select individual SharePoint
content databases for recovery.
To perform a recovery of one or more individual Microsoft Office SharePoint Server
2007 content databases:
1. Open NetWorker Module for Microsoft Applications on the web front-end that
was used to perform the backups of the SharePoint 2007 Writer.
2. In the navigation tree, expand the APPLICATIONS folder and select the
SharePoint Services Writer.
3. Expand the SharePoint Services Writer folder until you find a database folder
with an entry similar to the following example:
Services\Mars1\db_1
where Mars1 is the name of the server and db_1 is the name of the database.
4. Select the databases to restore.
specified hosts and start the recovery processes.
6. At the remote host:
a. Open NetWorker Module for Microsoft Applications.
b. In the navigation tree, locate and mark the items.
c. Restore the items.
7. Return to NetWorker Module for Microsoft Applications on the web front-end,
and click Continue in the Restore Remote Dependencies dialog box.
monitoring window.
Recovery of SharePoint Search/Content Indexes
To perform a full recovery of Microsoft Office SharePoint Server 2007:
1. Open NetWorker Module for Microsoft Applications.
SPSearch VSS Writer and OSearch VSS Writer.
specified hosts and start the recovery process.
151
4. Go to the remote host, and verify that the configuration database needs to be
recovered. If it needs to be recovered, then recover it.
5. Return to NetWorker Module for Microsoft Applications and click Continue in
the Restore Remote Dependencies dialog box.
monitoring window.
Rollback of SharePoint SQL databases
NMM does not support a SharePoint SQL rollback. SharePoint content databases are
backed up by NMM, but the master SQL database is not. The master database can be
backed up separately and then used in a SharePoint SQL rollback.
To back up SharePoint for rollback of SharePoint SQL databases, take a backup of the
master database each time a NMM SharePoint backup is taken. To perform a full
rollback, restore the backups created by the NMM SharePoint backup and then
restore the SharePoint SQL master database.
SharePoint 2003 backup and recovery
Microsoft SharePoint Portal Server 2003 and Windows SharePoint Server (WSS) 2.0
do not have their own VSS writers. Microsoft SQL Server hosts the SharePoint 2003
databases. NetWorker Module for Microsoft Applications uses Microsoft SQL writers
to perform the backup and recovery of the Microsoft SQL databases. After the
Microsoft SQL databases are recovered, the SharePoint recovery is performed.
NMM provides these backup features for SharePoint 2003 and where noted, in
Windows SharePoint Services 2.0:
◆
SharePoint 2003 database backup:
• Configuration database (also Windows SharePoint Services 2.0)
◆
SharePoint Portal Server 2003 portal sites backup, including the following
databases:
• Content database (also Windows SharePoint Services 2.0)
• User profile database
• Services database
• Index databases
• Windows SharePoint Services database backup
◆
Microsoft SQL Server-based Document and Picture Libraries (also Windows
SharePoint Services 2.0)
After a full backup of a database, individual items can be recovered from the
database.
◆
Web storage system-based Document Library Store and Document Libraries
After a full backup of a library, individual items can be recovered to the web
storage system-based library or redirected to filepaths.
◆
152
Backup level supported: Full only
Components used by NMM for SQL backup and recovery
NMM uses the following writers and commands to back up SharePoint 2003
databases hosted on Microsoft SQL Server:
◆
Application writers — NMM uses the application writers provided by Microsoft
SQL Server:
• MSDEWriter- SQL2000 — The writer for SQL 2000/MSDE
• SqlServerWriter — The writer for SQL Server 2005
“Specifying save sets for SharePoint 2003 SQL data” on page 153 describes
how the writer name is used in the save set syntax.
◆
Nsrsnap_vss_save — The NMM command to initiate the VSS-based backup
◆
Nsrsnap_vss_recover — The NMM command to initiate the VSS-based recovery
Specifying save sets for SharePoint 2003 SQL data
Table 26 on page 153 lists the SharePoint 2003 save set syntax to specify for supported
types of SQL data. Specify SQL data save sets in the Save Set attribute of a Client
resource.
Table 26
SQL save set syntax for SharePoint 2003 (page 1 of 2)
Save set syntax
databases including express and stand-alone
databases.
If the SQL Server VSS Writer service is disabled, SQL Server
2005 named instances and databases, including express and
stand-alone databases, can be backed up with the following
syntax:
APPLICATIONS:\MSDEWriter-SQL2000
153
Table 26
SQL save set syntax for SharePoint 2003 (page 2 of 2)
Save set syntax
database name combination.
nstance\master
For example, to back up the database named test11_SITE on
the SQL Server named instance GALL, type the following:
APPLICATIONS:\SqlServerWriter\GALL%5Ct
est11_SITE
values. “URL encoding for SQL save sets in SharePoint 2003
backup” on page 155 provides more information about URL
encoding.
All SQL 2000 named instances and
stand-alone databases.
database name combination.
APPLICATIONS:\MSDEWriter-SQL2000\host%
5C
instance\master
For example, to back up the database named test11_SERV on
the SQL Server named instance GALL, type the following:
APPLICATIONS:\MSDEWriter-SQL2000\GALL%
5Ctest11_SERV
database name must be replaced with URL-encoded values.
“URL encoding for SQL save sets in SharePoint 2003 backup”
on page 155 provides more information about URL encoding.
Displaying valid SQL data save sets for SharePoint 2003 backup
To display a list of the SQL save sets that are available for SharePoint 2003 backup:
nsrsnap_vss_save -?
3. Press Enter.
The save sets that are available for backup are listed in a format similar to the
following:
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServer
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServerTempDB
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CSPS01_Config_db
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_PROF
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SERV
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SITE
154
Set attribute of a Client resource. Each entry that you add to the Save Set
attribute must be typed on a separate line.
URL encoding for SQL save sets in SharePoint 2003 backup
◆
◆
two parts are always separated with a backslash (\) character. When specifying a
SQL Server named instance in the Save Set attribute, replace the backslash with its
URL-encoded value, which is %5C.
In the following example, a SQL Server named instance that is identified as:
GALL\test11_PROF
GALL%5Ctest11_PROF
To back up only the database named test11_PROF in the SQL Server named instance
GALL, the complete entry in the Save Set attribute would be:
APPLICATIONS:\SqlServerWriter\GALL%5Ctest11_PROF
If the database name in the previous example was test#11_PROF, the save set entry
would be:
APPLICATIONS:\SqlServerWriter\GALL%5Ctest%2311_PROF
values.
Table 27
Special character
URL-encoded value
Special character
URL-encoded value
\
%5C
?
%3F
/
%2F
]
%5D
"
%22
[
%5B
%
%25
}
%7D
#
%23
{
%7B
&
%26
^
%5E
<
%3C
‘
%60
>
%3E
|
%7C
155
SQL application information variable settings in SharePoint 2003 backup
Table 28
156
SQL variable settings for application information attribute
Attribute name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
application information
attribute
that moves snapshots from the
NMM to primary and secondary
(conventional) storage. The host
can be either a local host or a proxy
client host.
• The local host
client.
If setting up an SQL Client resource, use
specified, SQL consistency checks are
performed on the proxy client. Related
messages are written to the Replication
Manager log files (erm_clientxx.log) on
the proxy client.
If no proxy client is specified for an SQL
Client resource, consistency checks are
performed on the Client resource host.
Related messages are logged on the
Client resource host in both the
nmm.raw file.
SharePoint 2003 does not have its own VSS writers. To back up SharePoint 2003
databases follow the tasks for backing up a Microsoft SQL Server database.
The first four tasks for scheduling a backup are the same for all VSS writers
tasks in detail:
◆
◆
◆
◆
The next task is specific to configuring a SQL Server Client Resource for a SharePoint
database:
◆
“Task 5: Configuring a SharePoint 2003 Client resource” on page 157
Task 5: Configuring a SharePoint 2003 Client resource
The backup will use either SqlServerWriter or MSDEWriter-SQL2000, depending on
which writer is displayed when you run the nsrsnap_vss_save-? command to list the
available save sets. You can back up all SQL databases on the server, or individual
databases, depending on what you specify in the save set.
To configure a SharePoint 2003 Client resource using SQL Server writers:
Configuration.
The backup will use either SqlServerWriter or MSDEWriter-SQL2000, depending
on which writer is displayed when you run the nsrsnap_vss_save-? command to
list the available save sets. You can back up all SQL databases on the server, or
individual databases, depending on what you specify in the save set.
• To back up all SQL 2000 embedded databases, specify the save set name:
SYSTEM COMPONENTS:\
Note: The SYSTEM COMPONENTS:\ save set should be backed up as part the system
disaster recovery strategy.
157
• To back up all SQL 2000 named instances and stand-alone databases, specify
the save set name:
• To back up only the SQL 2000 host instance and database name combination,
specify the save set name:
APPLICATIONS:\MSDEWriter-SQL2000\host%5Cinstance\master
• To back up all SQL Server 2005 named instances and stand-alone databases,
If the SQL Server 2005 VSS Writer service is running, then SQL 2005 Express
databases are backed up under SYSTEM COMPONENTS:
SYSTEM COMPONENTS:\
If the SQL Server 2005 VSS Writer service is not running, then SQL 2005
Express databases are backed up under MSDEWriter-SQL2000:
The SQL Server save set syntax is described in “Specifying SharePoint 2007 save
sets for application data” on page 137.
backslash (\), and a table of values is provided in “URL encoding for SQL save
sets in SharePoint 2003 backup” on page 155.
10. Specify SQL application variable settings in the Application information
attribute of the Client resource. These settings are described in Table 28 on
page 156.
158
Performing a SharePoint 2003 recovery
The SQL Server databases used by SharePoint 2003 are recovered with NMM the
same as any other SQL Server databases. “Performing SQL Server recovery” on
page 121 provides the steps to recover the SQL Server databases to the SQL Server.
After recovering the databases, they may need to be updated and added to the
SharePoint portal.
To recover the SQL Server databases and restore them in SharePoint, use the method
that corresponds to the type of SQL database you are recovering:
◆
“Recovering SQL Server 2000 stand-alone databases for SharePoint 2003” on
page 159
◆
“Recovering SQL 2005 databases for SharePoint 2003” on page 159
Recovering SQL Server 2000 stand-alone databases for SharePoint 2003
To recover SQL Server 2000 stand-alone databases:
1. Stop the SQL Server services for the SQL Server 2000 stand-alone database
instances.
MSDEWriter-SQL2000 folder.
5. After the recovery is complete, start the SQL Server services for the SQL 2000
stand-alone database instances.
6. Launch SQL Enterprise Manager to verify that the databases have been
successfully recovered within SQL Server.
7. Launch SharePoint Server 2003 Central Administration.
8. Click List and Manage Portal Sites.
9. If the databases or portals are not displayed, click Restore Portal Site, and follow
the instructions to add the databases and restore the portal site.
Recovering SQL 2005 databases for SharePoint 2003
If the SQL Server VSS Writer service was enabled during backup, then SQL Server
2005 Express databases must be recovered from the SYSTEM COMPONENTS:\
save set.
If the SQL Server VSS Writer service was disabled during backup, then SQL Server
2005 Express databases must be recovered from the
APPLICATIONS:\MSDEWriter-SQL2000 save set.
The SQL Server VSS Writer service is disabled by default when Microsoft SQL Server
2005 is installed.
159
To recover SQL 2005 stand-alone databases that were backed up to the
SQLServerWriter save set:
1. If restoring the master database, stop the SQL Server services for the SQL 2005
database instances.
SQLServerWriter folder.
database instances.
To recover SQL 2005 express databases that were backed up to the SYSTEM
COMPONENTS save set:
If the SQL Server VSS Writer service was enabled during backup, then SQL 2005
Express databases must be recovered from the SYSTEM COMPONENTS:\ save set.
In this case, follow these steps:
1. If restoring the master database, stop the SQL Server services for the SQL 2005
database instances.
2. From the navigation tree, expand the SYSTEM COMPONENTS:\ folder.
database instances.
To recover SQL Server 2005 stand-alone or express databases that were backed up to
the MSDE-Writer-SQL2000 save set:
If the SQL Server VSS Writer service was disabled during backup, then SQL Server
2005 express and stand-alone databases must be recovered from the
MSDEWriter-SQL2000 save set. In this case, follow these steps:
1. Stop the SQL Server services for the SQL Server 2005 database instances.
MSDEWriter-SQL2000 folder.
160
5. After the recovery is complete, start the SQL Server services for the SQL Server
2005 database instances.
Best practices for SQL Server backup and recovery of SharePoint 2003 databases
Table 29 on page 161 lists the best practices and recommendations to follow when
using NMM to back up and recover SQL Server.
Table 29
SQL Server backup and recovery best practices and considerations (page 1 of 2)
Consideration
Best practice
For application servers such as SQL servers, consider backing up the
backup group.
application server
program
mirroring
Split-mirror technology is supported only with the EMC VSS Provider.
the NMM
If you attempt to use both the NMM and a NetWorker module such as
NetWorker Module for Microsoft SQL Server to back up application data,
the module backups will be promoted to full backups. “NMM client issues”
on page 340 provides more information about client issues.
Microsoft SQL data
If SQL 2005 backups were performed when the SQL Server VSS Writer
service was disabled, and you subsequently enable the writer, perform a
complete backup of SQL 2005 data. Performing a complete backup
enables you to recover all SQL 2005 data from the same application
folder, which is named SQLServerWriter.
161
Table 29
162
SQL Server backup and recovery best practices and considerations (page 2 of 2)
Consideration
Best practice
CLARiiON limits for the
CX500
If the number of snapshots exceeds the limit for number of snapshots
supported per CLARiiON, a "VETO" failure will occur during snapshot
creation. The CLARiiON limits for the CX500 are as follows:
For a CX500, the limit is 150 snapshots or snapshot sessions
system-wide. This includes reserved snapshots/sessions for SAN Copy.
There is also a limit of 8 snapshots or snapshot sessions per LUN.
Limits for CLARiiON systems can be found in the EMC SnapView
Integration Module for SQL Server (SIMS) Administrator’s Guide.
SQL instances in suspect
mode
If any of the databases belonging to a SQL instance are in suspect mode,
and a backup of the SQL instance is attempted with the Microsoft MSDE
Writer, the snapshot backup will fail.
Verify VSS SQL Writer
service is running
After installing the SQL Server, especially SQL 2005, a user needs to
verify that the VSS SQL Writer service is up and running in order for the
VSS SQL Writer snapshot to be successful. The SQL Writer is installed
but is disabled by default. A user must enable it or set it to automatic to
use its functionality.
Ignore tempdb database
Microsoft indicates that tempdb is a database that is rebuilt when the
server is rebooted, and that VSS Requestors can ignore this database.
The database is not reported by the writer and will not be backed up.
Special Character
databases
Special Character databases are not backed up and the best practice is to
have the list of databases without special character tables and columns.
Some SQL service may
have to be manually shut
down to perform
recovery
Administrators may have to manually shut down some SQL services to
restore SQL system databases. For SQL 2000, manually shut down the
SQL Server and the SQL Server Agent. For MSDE or SQL 2005, shut
down the "master" and "msdb" database only if it is being recovered.
Otherwise, do not shut down any MSDE or SQL 2005 database services.
SQL restore does not automate this process, and requires VSS
Requestors to take responsibility for controlling the SQL services during
restore.
SharePoint 2003 scenario of backup and recovery process
The backup and recovery for SharePoint 2003 and Windows SharePoint Services 2.0
configuration database is a backup of a Microsoft SQL Server database.
In this example, an organization is running Microsoft SharePoint Portal Server 2003
Service Pack 2/Windows SharePoint Services 2.0 Service Pack 2 on Windows Server
2003. It uses SQL 2000/SQL 2005 for the database server. The backup and recovery is
performed with NMM 2.2.
The following steps describe a backup and recovery operation performed by the
NMM administrator:
1. Creating a new SharePoint portal with three new databases.
2. Viewing the location and hierarchy of the databases as they appear in SQL
Enterprise Manager UI.
3. Configuring and performing a backup of the new databases with NMM savesets.
4. Verifying that all databases are backed up through the NMM UI.
5. Deleting the new databases.
6. Performing an NMM recovery of the databases.
7. Verifying that the databases have been recovered through the SQL Enterprise
Manager.
8. Recovering the portal site through SharePoint Portal Server Central
Administration.
9. Confirming the creation of the portal site in SharePoint Server Central
Administration.
Note: This descriptive example provides an overview of the interaction between Microsoft SQL
Server, NMM, and Microsoft SharePoint Server. For detailed information on using Microsoft
SQL Server and Microsoft SharePoint Server, consult the respective Microsoft documentation
for those products.
Creating a new SharePoint portal with three new databases
The SharePoint administrator uses SharePoint Portal Server Central Administration
to create a new portal “test11” with three SQL databases:
◆
test11_SITE
◆
test11_PROF
◆
test11_SERV
The organization puts the portal into use by creating sites, and by uploading
documents, pictures, events, and other items.
163
Viewing the databases as they appear in SQL Enterprise Manager UI
Figure 17 on page 164 displays the new databases as they appear in SQL Enterprise
Manager.
Figure 17
Newly created databases created in SharePoint
Configuring and performing a backup of the new databases with NMM save sets
The administrator sets up a backup of the SQL database. The administrator runs the
following command on a command line:
nsrsnap_vss_save -?
This command lists the SQL writer, and the save sets available for backup.
“Displaying valid SQL data save sets” on page 116 provides more information about
using this command. The syntax for save sets is described in detail in “URL encoding
for SQL save sets” on page 117. In this example, the databases created by the
organization are as follows:
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServer
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServerTempDB
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CSPS01_Config_db
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_PROF
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SERV
APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SITE
The administrator determined the save sets available, and followed the steps for
creating a SQL Server backup as described in “Performing SQL Server backups” on
page 119. The administrator specified the save set
“APPLICATIONS:\MSDEWriter-SQL2000” to back up all of the SharePoint
databases. To back up just the “test11_PROF” database, the administrator would
specify “APPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_PROF.”
Verifying that all databases are backed up through the UI
The administrator verified that the databases had been backed up by looking at the
available recovery sets in NMM. In the navigation tree, the administrator expanded
the Applications folder, the MSDEWriter-SQL2000 folder, and then the GALL
(server) folder. Figure 18 on page 165 shows that all of the databases, including the
three new “test11...” databases, were backed up.
164
Figure 18
NMM backups available for recovery
Deleting the new databases
The administrator wants to test NMM recovery, and deletes the portal from the
SharePoint Server. This deletion of the portal deletes the three new “test11...”
databases from SQL. During deletion, SharePoint displays a message to verify that
the administrator wants to delete the databases. The administrator clicks Yes and the
databases are deleted.
Performing an NMM recovery of the databases
The administrator opens NMM again to recover the three databases. The
administrator selects the new databases that were just deleted from SharePoint.
Figure 19 on page 165 shows test11_PROF, test11_SERV, and test11_SITE databases
selected by the administrator.
Figure 19
SQL databases selected for recovery in NMM
The administrator clicks Start Recover. The recovery is successful.
165
Verifying database recovery through the SQL Enterprise Manager
The administrator launches SQL Enterprise Manager and verifies that all three
databases have been recovered. Figure 20 on page 166 shows that the three databases
were recovered to the SQL Server.
Figure 20
Databases recovered to SQL Server
Recovering the portal site
Recovering the SQL databases does not automatically bring them back into
SharePoint. In SharePoint, the administrator launches SharePoint Portal Server
Central Administration. Figure 21 on page 166 shows the Portal Site and Virtual
Server Configuration window.
Figure 21
166
SharePoint Server Central Administration window
The administrator clicks List and manage portal sites. In Figure 22 on page 167, the
Manage Portal Sites window does not display the portal site.
Figure 22
SharePoint Portal Server Central Administration, Manage Portal Sites page
The administrator clicks Recover Portal Site. Figure 23 on page 167 shows the options
for creating and restoring the portal.
Figure 23
Portal creation and restore options
167
Figure 24 on page 168 shows the settings specified by the administrator: Restore a
portal, and the Site database name “test11_SITE,” User profile database name
“test11_PROF,” and Services database name “test11_SERV.”
Figure 24
Portal restore and database names
After completing this page, a confirmation page is displayed, as shown in Figure 25
on page 168.
Figure 25
168
Confirmation page for request to create a portal site
The administrator clicks OK and portal creation begins. Figure 26 on page 169 shows
the status page that is displayed in SharePoint Portal Server Central Administration
during portal creation.
Figure 26
Status page for portal creation process
Figure 27 on page 169 shows the page displayed to indicate that the portal creation
was successful.
Figure 27
Operation Successful page displayed when portal creation is complete
Verifying the creation of the portal site in SharePoint Server Central Administration
In SharePoint Portal Server Central Administration, the administrator launches the
Manage Portal Sites page, and the Portal Site is now listed as shown in Figure 28 on
page 169.
Figure 28
Recovered databases and SharePoint portal visible and available in SharePoint
169
SharePoint 2007 disaster recovery
One of the keys to successful disaster recovery is making sure you have a full and
complete backup available before disaster strikes. Plan your backup schedule and
strategy to ensure that you always have an up-to-date backup available in case of
disaster.
SharePoint Disaster Recovery is more than restoring a full backup. NMM backs up
the SharePoint configuration and content databases, and the query and index servers.
In a disaster recovery, you need to restore the Internet Information Service (IIS), SQL
Servers, and SharePoint Servers that make up the farm, including file systems and
registries of each machine.
Disaster recovery planning
The following sections describe the steps you must take as part of your disaster
recovery plan.
SharePoint Server backup list
Table 30 on page 170 lists each server type in the SharePoint farm and the items that
should be backed up on that server as part of your disaster recovery plan.
Table 30
SharePoint Server backup list for disaster recovery
Server type
Items to back up
Web Front-End
SYSTEM COMPONENTS:\
All System Volumes. For example, C:\, D:\,
APPLICATIONS:\Microsoft Office SharePoint Services\
Web Front-End without Microsoft Office
SharePoint Services Writer
SYSTEM COMPONENTS:\
All System Volumes For example, C:\, D:\,
Search Server
SYSTEM COMPONENTS:\
All System Volumes For example, C:\, D:\,.
SQL Server
SYSTEM COMPONENTS:\
All System Volumes For example, C:\, D:\,
All SQL databases, including the "master" database
Stand-alone SharePoint Server
SYSTEM COMPONENTS:\
All System Volumes (For example, C:\, D:\,
All system databases, including the "master" database
APPLICATIONS:\Microsoft Office SharePoint Services\
Additional requirements for successful SharePoint disaster recovery
The following conditions and requirements must be met for a successful disaster
recovery:
170
◆
The SYSTEM COMPONENTS:\ save set and all system volumes must be
recovered at the same time to ensure consistency on the system.
◆
For systems with remote dependencies, the remote machines must be recovered
first. Once all remote machines have been recovered, the primary system can then
be recovered.
◆
The recovery of the "master" database requires that SQL services are not running
at the time of recovery. If SQL services are running and an attempt to recover this
database is made, the recovery attempt will fail.
◆
The recovery of:
• SharePoint Servers that were backed up by NMM as described in “Disaster
• SharePoint farm as described in “Full recovery of a stand-alone SharePoint
farm” on page 149.
Backing up SharePoint critical components
The Microsoft Office SharePoint Server 2007 installation directory, web parts, IIS
virtual servers, <drive>:\Inetpub, and the Registry reside on all web front-end
servers, application servers, and database server. To protect these critical
components, ensure that you have a backup strategy that includes file system and
system state backups of each Sharepoint Server.
Back up the following folders by using NMM:
◆
Microsoft Office SharePoint Server 2007 installation directory on all the servers in
the farm
◆
SQL installation directory on database server
◆
Virtual directories of following IIS websites on all servers in the farm:
• Default website
• SharePoint Central Admin
• Web application
• SSP web application
◆
IIS, Microsoft Office SharePoint Server 2007, and SQL logs:
• Internet Information Services logs (*.log) at
C:\Winnt\System32\Logfiles\W3svc1\
• Usage analysis log at C:\Windows\System32\LogFiles\STS\ if usage
analysis is enabled
• Other Windows SharePoint Services logs, STSAdm.log and OWSTimer.log,
from the C:\Documents and
Settings\Windows_SharePoint_Services_Administrator_Account\Local
Settings\Temp directory
◆
SharePoint configuration files. The directory at C:\Program Files\Common
Files\Microsoft Shared\web server extensions
◆
IIS backup
◆
Registry backup on all servers
Backing up the Global Assembly Cache
The .NET assemblies installed in Global Assembly Cache (GAC) are not backed up by
any of the SharePoint writers. To back up the GAC, back up the items as regular file
system objects.
The DLLs located in GAC are installed by applications and can be reinstalled during
Disaster Recovery. The GAC includes SharePoint assemblies, and possibly your own
custom assemblies.
The GAC files are represented in Windows Explorer under
C:\WINDOWS\assembly.
SharePoint 2007 disaster recovery
171
The actual location can only be navigated to by using the command line. You cannot
use GUI navigation such as Windows Explorer.
For example, Microsoft.SharePoint.dll assembly is located in the following folder:
C:\WINDOWS\assembly\GAC_MSIL\Microsoft.SharePoint\12.0.0.0__71e9bce111e
9429c
Performing a disaster recovery
To perform a disaster recovery:
1. Back up the following:
• APPLICATIONS:\Microsoft Office SharePoint Services
• APPLICATIONS:\SqlServerWriter
• C:\
• SYSTEM COMPONENTS:\
In addition, you may want to back up various servers, folders, and other
components as described in “Disaster recovery planning” on page 170.
2. Prepare a setup similar to the one currently being used. Use the same OS, service
packs, IP address, hostname, and domain.
3. Install the same versions of SQL Server and SharePoint Server as currently being
used on the new setup.
4. First recover c:\, and SYSTEM COMPONENTS:\, and then the SQL data.
5. Perform a full recovery of the SharePoint writer.
“Performing SQL Server disaster recovery” on page 123 provides information for
disaster recovery of SQL Server on Windows Server.
!
IMPORTANT
Make sure to restart the system after performing a disaster recovery of SharePoint.
Currently a message about the completion of disaster recovery and the need to
restart the system appears in the Monitor window and the logs. Check the Monitor
window and the logs before restarting the system.
172
7
Microsoft SharePoint
2007 Granular Backup
and Recovery
◆
◆
◆
◆
◆
◆
◆
◆
◆
◆
◆
NMM support for granular backup and recovery of SharePoint items..............
Better performance for large SharePoint farms ......................................................
Performing SharePoint 2007 granular backups ......................................................
Configuring a SharePoint 2007 Client resource for granular backup..................
SharePoint 2007 granular recovery...........................................................................
SharePoint recovery search........................................................................................
Performing a granular recovery of SharePoint items ............................................
SharePoint prerecovery options ................................................................................
SharePoint recovery options......................................................................................
Commands for SharePoint 2007 granular backup and recovery .........................
Using administrator account other than that created during installation ..........
174
175
175
176
177
179
182
183
185
186
192
173
NMM support for granular backup and recovery of SharePoint items
NMM supports granular backup of SharePoint 2007 objects and granular recovery of
SharePoint 2007 items.
SharePoint objects available in NMM granular backup
Granular backup of SharePoint 2007 uses a different save set syntax than that used for
optimized SharePoint 2007 backup. Displaying valid SharePoint 2007 save sets for
granular recovery describes how to view and use this syntax to configure a Client
resource for NMM backup.
The following SharePoint objects are available for NMM granular backup:
◆
Farm
◆
Application
◆
Site collection
◆
Site
◆
Sub-site
◆
List
SharePoint limitations on document metadata recovery
Due to limitations in the Microsoft SharePoint APIs, some metadata associated with a
document may not be restored when an individual document is recovered in a
granular recovery. The document itself is recovered successfully, but some metadata
changes that occurred after the last successful data may be lost.
This is most likely to occur when a document is deleted from SharePoint, and then it
is retrieved in a granular recovery.
The metadata that is not recovered is not critical. However, there are several methods
for recovering the metadata when the document is recovered:
◆
Recover the entire top level website with NMM, and then retrieve the document
from that recovery. This will restore all content in the website, so all documents
and data on that website will be restored to the state they were in at the time of
the last backup.
or
◆
174
Within the SharePoint UI, retrieve the lost document from the SharePoint recycle
bin.
Better performance for large SharePoint farms
NMM release 2.2 provides better granular and optimized backup performance for
large SharePoint farms:
◆
Parallel staging for multiple site collections at once — You can run multiple save
sets in parallel, thereby increasing backup throughput and reducing backup time.
The number of parallel staging operations is limited to the number of processors
on the web front-end machine.
Note: A processor or CPU is an electronic circuit that can execute the programs. System
environment variable 'NUMBER_OF_PROCESSORS' stores the number of processors on a
current machine. For example, a quad-core processor is considered four, and four staging
threads run concurrently.
◆
Parallel backups from staging area to NetWorker Server for multiple
site-collections — The number of parallel backups to NetWorker Server is
controlled by parallelism set in NetWorker Client Configuration, when -P is
passed to nsr_moss_save.exe process.
◆
Running multiple save sets on a single web front-end.
Performing SharePoint 2007 granular backups
A SharePoint 2007 granular backup does not use the SQL and SharePoint VSS writers,
so it is not necessary to register these writers prior to creating a Client resource for
SharePoint 2007 granular backup.
The first four tasks for scheduling a backup are the same for all applications and VSS
writers supported by NMM. Chapter 2, “Configuring a Scheduled Backup,”
describes these tasks in detail:
◆
◆
◆
◆
The next task is specific to configuring a SharePoint 2007 Client resource for a
granular backup:
“Configuring a SharePoint 2007 Client resource for granular backup” on page 176
Better performance for large SharePoint farms
175
Configuring a SharePoint 2007 Client resource for granular backup
To configure a SharePoint 2007 Client resource for granular backup:
Configuration.
To back up the entire SharePoint farm, specify the save set name:
NMMOSS:/<FARM_NAME>
To back up a single web application, specify the save set name:
NMMOSS:/<FARM_NAME>/<WEB_APPLICATION>
To back up a SharePoint object by url, such as a site-collection, site, or sub-site,
NMMOSS:/URL=<url>
The SharePoint Server save set syntax is described in “Specifying save sets for
SharePoint granular backup” on page 186.
will be added.
nsr_moss_save.exe
OPTIMIZED=false
176
18. Click OK.
SharePoint 2007 granular recovery
You can individually select SharePoint items for recovery, from a granular
SharePoint 2007 object backup. The items can be located and selected by browsing the
navigation tree, or by searching for specific items.
NMM supports granular recovery across a wide variety of items and objects,
including:
◆
List/Library:
• Document Library
• Form Library
• Picture Library
• Import Spreadsheet
• Announcements
• Contacts
• Discussion Board
• Links
• GenericList
• XMLForm
• Calendar
• Tasks
• Issue Tracking
• Survey
• List Template Gallery
• Site Template Gallery
• Reporting Metadata
• Custom List in Data Sheet View
• Blogs
• Wiki Page Library
• Custom List
• Translation Management Library
• Data Connection Library
• Slide Library
• Report Library
• Project Tasks
SharePoint 2007 granular recovery
177
• Languages and Translators
◆
Site-Collection Template:
• Team Site
• Blank Site
• Document Workspace
• Basic Meeting Workspace
• Blank Meeting Workspace
• Decision Meeting Workspace
• Social Meeting Workspace
• MultiPage Meeting Workspace
• Document Center
• Record Center
• Personalization Site
• Site Directory
• Report Center
• My Site Host
• Collaboration Portal
• Publishing Portal
• News Site
• Publishing Site
• Publishing Site With WorkFlow
• Wiki Site
• Blog Site
• Search Center with Tabs
• Search Center
◆
Miscellaneous:
• SharePoint Farms
• Sites and Workspaces
• Web Applications
• Site Collections
• Sites
To locate SharePoint items to recover, NMM provides a navigation tree and advanced
search features for SharePoint. “SharePoint recovery search” on page 179 describes
how to use the SharePoint recovery search features.
“Performing a granular recovery of SharePoint items” on page 182 provides detailed
information on how to recover SharePoint items.
178
SharePoint recovery search
NMM provides two methods for locating and selecting SharePoint items for recovery:
◆
Browsing through the navigation tree
◆
Searching the NetWorker index
To support locating and selecting individual items for recovery, NMM’s SharePoint
advanced search allows you to search backup data by:
◆
Document Name
◆
Document Type
◆
Backup Time
◆
Creation Date
◆
Modification Date
◆
Author
◆
Minimum Size
◆
Document Title
The items available for recovery depend on the actual items contained in your
specific databases.
Figure 29 on page 179 provides an example of how items are displayed in the NMM
Recover window. You can browse the navigation tree and expand folders. Folders
can be selected for recovery in the navigation tree, or individual items can be
recovered in the item list.
Figure 29
NMM displaying SharePoint 2007 browsing for recovery items
179
If you are looking for a particular item it may be easier to locate the item by searching,
instead of browsing folders until you find it. Figure 30 on page 180 shows the search
fields available on the Search tab. Table 31 on page 181 describes the search fields that
are available.
Figure 30
180
NMM displaying Search tab for SharePoint 2007 recovery
Table 31
SharePoint recovery search fields
Field
Description
Name
The search criteria. What you are searching for.
Type
Select an object type from the list of supported SharePoint 2007 object types.
Backup Time
Select the type of range, Before, Between, or After and then specify the date or date
range. If not selected, the default for this field is Before and a range of 90 days back
from the current date.
Creation Date
range. If not selected, this field is ignored.
Modification Date
range. If not selected, this field is ignored.
Author
Type the name of the author of the item. This search field is usually associated with
list item objects, files, and documents.
Minimum Size
Type a number that represents the minimum size to search for in kilobytes (Kb),
megabytes (Mb) or gigabytes (Gb) depending on the selection. This field only
accepts integers. Alphabetical or special characters are not valid entries.
Title
Type the title of the object. This field is usually associated with list item objects.
Searching and marking tips
There are several ways you can improve your searching and marking:
◆
Each search criteria is treated as an AND item in the search query. Search will
only return items that match all criteria specified. For example, if you specify to
search for an Object Name “Status.doc” and Minimum Size “150 Kb,” search will
only return items that match all of the criteria: Status.doc and at least 150 Kb. It
will not return any items named Status.doc that are less than 150 Kb, or any other
.doc files that are at least 150 Kb.
◆
Marked search items will not appear in the task pane. For “list item” object type,
multiple items may be marked and restored. Multiple versions of the same item
may not be marked. For all other object type searches, only a single item may be
marked at a time.
◆
The maximum number of search results returned is 5000. If the search item does
not appear in the results, reduce the Backup Time range.
Searching for a SharePoint item
To search for a SharePoint item:
1. Start the NMM Client program.
2. From the application toolbar, click the NetWorker Server icon to select the
NetWorker server on which the NetWorker client was configured for backup.
3. In the left pane, select SharePoint Recovery.
4. Click the Browse tab.
5. Navigate to the level in the tree you want to search from, or select root if you do
not know where the item resides.
181
6. Right-click the node in the browse tree, and then click Search.
7. Select the checkbox for each search type to include in the search.
8. Specify what to search for in each field.
9. In the Search Results pane, mark the item to restore.
10. Select recover options.
11. Click Start Recovery.
Performing a granular recovery of SharePoint items
To recover a SharePoint object:
3. In the left pane, select SharePoint Recovery.
4. In the navigation tree, select the SharePoint objects to be recovered. By default,
the objects displayed in the navigation tree are from the most recent backup.
To recover objects from a previous backup:
• To view all versions of a backup object prior to the selected browse time, select
an object in the navigation tree, right-click and select Versions.
• To search for specific items, click the Search tab.
5. Locate and select the items for recovery. To view more information about a
SharePoint object, right-click the object and then click Properties.
6. From the Advanced Recover toolbar, click Start Recover.
The Pre-Recovery Options dialog box appears. This dialog box specifies where to
place the restored items, how to update items, and the security permissions
required. “SharePoint prerecovery options” on page 183 describes these options
in detail.
7. Specify the options in Pre-Recovery Options and click OK.
The SharePoint Recover Options dialog box appears.
This dialog box specifies where to restore items from, the location of the restore
log folder, and how to lock the files after the restore job. “SharePoint recovery
options” on page 185 describes these options in detail.
8. Specify the options in the SharePoint Recover Options and then click OK.
182
SharePoint prerecovery options
The Pre-Recovery Options dialog box specifies where to place the restored items, how
to update items, and the security permissions required. “Performing a granular
recovery of SharePoint items” on page 182 describes the recovery process and steps
in detail.
The default settings for this dialog box are:
◆
Recover Type: Recover to original location
◆
Update Version: Add as a new version
◆
Security Permissions: Inherit from parent
Selecting alternate destination options
For SharePoint, NMM provides directed recovery through restore to an alternate location.
You can choose to restore objects to an alternate web application. The alternate
location can be within the same farm as the backed up object or to an alternate farm.
Restore to an alternate location is useful for the following:
◆
Back up site collection and restore to an alternate web application.
◆
Back up website and restore to alternate site-collection or sub-site.
◆
Back up list and restore to subsite or sub-web or root-web.
◆
Back up list items and restore to same list and same hierarchy in alternate
location.
To specify Pre-Recovery Options:
1. Specify the Recovery Type.
If Restore to Original Location is selected, proceed to step 4 on page 184.
If Restore to Alternate Location is selected, proceed to step 3 on page 183.
2. Click Refresh to make sure that all alternate locations in the same or alternate
farm appear in the lists that are provided.
3. Select the alternate location from the lists:
The Farm list contains all the available farms. The farm may be the same as the
backed up object or an alternate farm.
The Web Front End list contains all the web front-end machines that constitutes
the selected Farm. One SharePoint farm may constitute of many web front-ends,
such as application sever, database server, index server, and normal web
front-ends.
The Web Applications list contains all of the web applications available in this
SharePoint farm. If restoring a site object, only the web applications must be
selected.
The Site Collection list contains all of the site collections available from the
selected web application. if restoring a web object, both the web application and
site-collection must be specified.
The Web Site list contains all of the websites available from the selected Site
Collection. If restoring a list object, the web applications, site-collection, and
website must be specified.
SharePoint prerecovery options
183
4. The options under Conflict Resolution specify how to handle different versions
when there are conflicts between the recovery version and the current version in
the selected original or alternate location:
• Select Restore over existing items to overwritten the existing version in the
original or alternate location with the recovered version. After recovery, only
the recovery version is available.
• Select Skip if the item exists to retain the existing version in the original or
alternate location, and ignore the recovery version if it is different from the
existing version.
• Select Add as a new version to add the recovered version in the original or
alternate location. After recovery, the current version and the recovered
version are both available. If version control is not enabled, then the existing
version is overwritten by the new version.
5. Select one of the following options under Security Permissions:
• Inherit from parent (Default) to use the SharePoint security permissions of the
parent container.
• WSS Only to include user memberships, role assignments such as
out-of-the-box roles like Web Designer, or any custom roles that extend from
the out-of-the-box roles. The ACL for each object is restored. No user
information defined in the DAP or LDAP servers is included.
• All to include user memberships, role assignments such as out-of-the-box
roles like Web Designer, or any custom roles that extend from the
out-of-the-box roles. The ACL for each object is restored. In addition, user
information defined in the DAP or LDAP servers is included.
The SharePoint Configuration Options dialog box appears. “SharePoint
recovery options” on page 185 describes in detail how to specify these settings.
184
SharePoint recovery options
The SharePoint Recovery Option dialog box specifies for all recovery jobs where to
restore items from, the location of the Restore Log folder, and how to lock the files
after the restore job. These settings are applied to every restore job. “Performing a
granular recovery of SharePoint items” on page 182 describes the recovery process
and steps in detail.
Once a restore is complete, NMM automatically enables SharePoint write-protect.
This allows you to review the restored SharePoint Object before allowing other users
to visit and make changes to the SharePoint Object. If a user attempts to use the
website while the write-protect is in place, the following message appears,
“Additions to this website have been blocked. Contact the administrator to resolve
this problem. Troubleshoot issues with Windows SharePoint Services.”
To specify recovery options:
1. In the Restore From box, type the file path or click the Browse button to specify
the temporary staging area location.
The Restore From location is a temporary staging area for the data from the
NetWorker Server. NetWorker SharePoint Service Controller service picks up this
data and sends it to the SharePoint databases. The staging area space is reclaimed
after recovery.
2. In the Restore Log Folder, type the file path where the restore log file should be
placed.
When NMM restores SharePoint Objects, a restore log file is created during the
restore.
3. In the Lock - Apply lock after restore job group, specify the read and write lock
settings for the items after restore is complete.
4. Specify the SharePoint Configuration options and then click OK.
SharePoint recovery options
185
Commands for SharePoint 2007 granular backup and recovery
NMM uses special commands for SharePoint 2007 granular backup and recovery.
nsr_moss_save.exe command
The nsr_moss_save.exe command is used to query or get the save set syntax of
SharePoint objects. You can configure this command in the Backup Command field of
the Client resource in the NMC. Once the save set is retrieved, use the save set in the
Save Set field of the Client resource in the NMC. Multiple save sets commands are
supported in this release.
To display information about SharePoint 2007 granular backup options, operations,
and parameters, at the command line type:
nsr_moss_save.exe -?
This command can be used to display save set syntax for SharePoint farm name, web
applications, site-collections, sub-site collections, and list.
nsr_moss_save.exe [-s server] [-c client] [-N saveset] [-b pool] [-g group] [-l
(full | incr)] -A OPTIMIZED=false/true
Specifying save sets for SharePoint granular backup
Table 32 on page 186 lists the SharePoint save set syntax to specify for granular
backup of SharePoint objects. Specify SharePoint object save sets in the Save Set
attribute of a Client resource.
Table 32
186
SharePoint granular recovery save set syntax for SharePoint 2007 objects
Save set syntax
Entire SharePoint farm
NMMOSS:/ or NMMOSS:/<FARM_NAME>
Single web application
NMMOSS:/<FARM_NAME>/<WEB_APPLICATION>
A SharePoint object specified by URL, such
as a site-collection, site, or sub-site
NMMOSS:/URL=<url>
A SharePoint list specified by URL and list
name
NMMOSS:/Url=<url>;Name=/<List
Name>;Type=List
Displaying valid SharePoint 2007 syntax for granular backup
The nsr_moss_save command is used to display valid save sets that are available for
granular backup.
Figure 31 on page 187 displays the output of the nsr_moss_save command when
used with several different switches.
Figure 31
Sample output from nsr_moss_save
The following example shows the Name and Save Set Syntax displayed at the
command line. The save set syntax for a SharePoint object may include long names,
so when you run nsr_moss_save command, the output may wrap to more than one
line in the command window for each save set.
nsr_moss_save -f
Name
SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569
NMMOSS:/SharePoint_Config_5e
687d39-3cf4-4769-b88b-1d3ead776569
Save Set Syntax
The save set syntax from the previous example is the following string:
NMMOSS:/SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569
Note: When copying Save Set Syntax from a command window, be sure to get the entire string.
Always use the full Save Set Syntax for backups.
187
To display a list of the SharePoint 2007 save sets that are available for granular
backup:
2. From the command prompt, type one of the following commands to list the valid
SharePoint objects available:
• nsr_moss_save [-f]
Displays the current SharePoint farm name and save set syntax.
The save set syntax is the type of data to back up combined with the object
name. For example:
nsr_moss_save -f
Name
Save Set Syntax
SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569NMMOSS:/SharePoint_Config_5e
687d39-3cf4-4769-b88b-1d3ead776569
• nsr_moss_save [-w]
Displays all web applications and save set syntax. For example:
nsr_moss_save -w
All Web Applications:
Name
Save Set Syntax
SharePoint - 80
687d39-3cf4-4769-b88b-1d3ead776569/SharePoint - 80
@CentralAdmin
687d39-3cf4-4769-b88b-1d3ead776569/@CentralAdmin
SharePoint - 26061
687d39-3cf4-4769-b88b-1d3ead776569/SharePoint - 26061
• nsr_moss_save [-w webapplication]
Displays all site-collections of a web application and save set syntax. For
example, the site-collections in the web application SharePoint - 80:
nsr_moss_save -w “SharePoint - 80”
All site-collections for Web Application: SharePoint - 80
Name
Save Set Syntax
http://mj-hvmoss
NMMOSS:/Url=http:/mj-hvmoss
http://mj-hvmoss/sites/Sunil
NMMOSS:/Url=http://mj-hvmoss/sites/Sunil
http://mj-hvmoss/MySite
NMMOSS:/Url=http://mj-hvmoss/MySite
• nsr_moss_save [-u site-collectionl url]
Displays all sub-sites of a site-collection and save set syntax. For example, the
sub-sites for the site-collection http://inenbeherad2c:48137/sites/dsd:
nsr_moss_save.exe -u http://inenbeherad2c:48137/sites/dsd
All Sites for Site-Collection: http://inenbeherad2c:48137/sites/dsd
Name
Saveset Syntax
http://inenbeherad2c:48137/sites/dsd/subsite
NMMOSS:/Url=http://
inenbeherad2c:48137/sites/dsd/subsite
http://inenbeherad2c:48137/sites/dsd
NMMOSS:/Url=http://
inenbeherad2c:48137/sites/dsd
http://inenbeherad2c:48137/sites/dsd/subsite/subsite
NMMOSS:/Url=http://
inenbeherad2c:48137/sites/dsd/subsite/subsite
188
• nsr_moss_save [-k website or sub-web url]
Displays all the lists of specified websites or sub-web. For example, the
sub-site for website http://inenbeherad2c:48137/sites/dsd:
nsr_moss_save.exe -k http://inenbeherad2c:48137/sites/dsd
All Lists for Web: http://inenbeherad2c:48137/sites/dsd
Name
Saveset Syntax
Links
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Links;Type=List
Tasks
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Tasks;Type=List
List Template Gallery
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/List Template
Gallery;Type=List Shared Documents
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Shared Documents;Type=List
Form Templates
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Form Templates;Type=List
Reporting Metadata
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Reporting Metadata;Type=List
Converted Forms
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Converted Forms;Type=List
Team Discussion
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Team Discussion;Type=List
Calendar
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Calendar;Type=List
Master Page Gallery
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Master Page Gallery;Type=List
Announcements
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Announcements;Type=List
Reporting Templates
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Reporting
Templates;Type=List
Web Part Gallery
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Web Part Gallery;Type=List
User Information List
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/User Information
List;Type=List
Site Template Gallery
NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Site Template
Gallery;Type=List
Each line of output provides a save set entry that you can add to the Save Set
attribute of a Client resource. Each entry that you add to the Save Set attribute must
be typed on a separate line.
SharePoint 2007 granular application information variable settings
Table 33 on page 189 describes the variable that can be specified in the application
Table 33
Attribute name
Description
Values
-A OPTIMIZED=value
For granular backup use
-A OPTIMIZED=false.
The default value is false.
true | false
For example,
nsr_moss_save.exe –s
<server> -c <client> -A
OPTIMIZED=false
189
nsr_moss_recover.exe command
The nsr_moss_recover.exe command is used at the command line to display
SharePoint 2007 granular recovery options. The syntax for specifying SharePoint 2007
granular recovery save sets is different from the VSS writer syntax used for other
applications in NMM.
To display information about SharePoint 2007 granular recovery options, operations,
and parameters, at the command line type:
nsr_moss_recover.exe -?
Table 34 on page 190 lists the options that can be used with nsr_moss_recover.exe
command.
Table 34
190
nsr_moss_recover command options
Command option
Description
site-url
URL of the site-collection
web-application
Name of the destination web application in the case of recovery to alternate location
web-url
URL of the website
web-guid
Guid of the website
retain-identity
0(default) - No, 1 - Yes
update-versions
1(defualt) - Append, 2 - Overwrite, 3 - Ignore, 4 - Error
update-user
0(default) - None, 1 - Replace, 2 - ImportAll
update-security
0(default) - None, 1 - WssOnly, 2 - All
halton-warning
halton-nonfatal
apply-locks
0(default) - Read Unlocked/Write Unlocked, 1 - Read Unlocked/Write Locked, 2 Read Locked/Write Unlocked, 3 - Read Locked/Write Locked
Examples of how to use the nsr_moss_recover.exe command:
nsr_moss_recover.exe [-fnqu] [-i {nNyYrR}] [-s server] [-c
client] [-t <date | "l locale_date">]
[-d staging-location] [-W web-application] [-T site-url] [-w
web-url] [-g web-guid] [-O retain-identity <0|1>]
[-V update-versions <1|2|3|4>] [-U update-user <0|1|2>] [-S
update-security <0|1|2>] [-E halt-onwarning <0|1>]
[-F halt-onnonfatal <0|1>] [-l apply-locks <0|1|2|3>] [dir]
or
nsr_moss_recover.exe [-fnqu] [-i {nNyYrR}] [-I input-file] [-s
server] [-c client] [-t <date | "l locale_date">]
[-d staging-location] [-W web-application] [-T site-url] [-w
web-url] [-g web-guid] [-O retain-identity <0|1>]
[-V update-versions <1|2|3|4>] [-U update-user <0|1|2>] [-S
update-security <0|1|2>] [-E halt-onwarning <0|1>]
[-F halt-onnonfatal <0|1>] [-l apply-locks <0|1|2|3>] [path...]
191
Using administrator account other than that created during installation
To run SharePoint granular backup and recoveries in NMM using an administrator
account other than that created during NMM installation, the following permissions
are required:
192
◆
Make user the domain user.
◆
Add user to SharePoint Farm Administrators list giving full control.
◆
Add user to the list of site administrators for each site collection in the farm.
◆
Using SQL Server Enterprise manager, assign db_creator, public, and security admin
server roles to the user.
◆
Using SQL Server Enterprise manager, assign db_owner and public user mappings
for each databases in the SharePoint farm.
8
Microsoft Exchange
Server Backup and
Recovery
◆
◆
◆
◆
◆
◆
◆
Exchange Server backup and recovery ....................................................................
Performing Exchange Server backups .....................................................................
Configuring an Exchange Client resource...............................................................
Performing Exchange Server recovery.....................................................................
Recovery to storage groups .......................................................................................
Microsoft Exchange Server 2007 directed recovery ...............................................
Microsoft Exchange Server 2007 disaster recovery ................................................
194
206
208
219
226
238
241
193
Exchange Server backup and recovery
(NMM) backup procedures in Chapter 2, “Configuring a Scheduled Backup.”
Microsoft Exchange Server 2003 and Exchange Server 2007 support Volume Shadow
Copy Service (VSS) through Microsoft-supplied application writers. Using the VSS
writers, you can back up and restore Exchange databases and transaction log files.
NMM can restore storage groups to the original Exchange Server and Storage Group,
to an alternate Storage Group on the same Exchange Server, to the Exchange
Recovery Storage Group (RSG) or to an alternate Exchange Server. NMM can also
recover Exchange mailboxes and email messages quickly by allowing Exchange RSG
recoveries and integrated RSG browsing for recovery.
NMM allows the creation of a point-in-time copy of data. The snapshot includes exact
copies of data and open files. This allows the backup of the Exchange server while the
Exchange database is online or in use.
Microsoft Exchange Server MAPI client and Collaboration Data Objects support
Access to Exchange messaging stores utilized by NMM mailbox backups requires the
Messaging API (MAPI) and Collaboration Data Objects (CDO) kit. The MAPI client
libraries and CDO are not included with Exchange Server 2007.
The package with these objects can be downloaded from Microsoft’s website. This
package is also required for RSG support in Exchange Server 2007. Download version
1.2.1 or later. For information on downloading and installing this package, search for
"Microsoft Exchange Server MAPI Client and Collaboration Data Objects" on the
Microsoft Download Center website.
Exchange Server versions supported by NMM client
NMM client supports the following versions of Microsoft Exchange Server:
◆
Microsoft Exchange Server 2003 (x86) SP2
◆
Microsoft Exchange Server 2007 (x64) SP1 or later
Note: Restore to Recovery Storage Group or alternate storage group is only available with
Microsoft Exchange Server 2007.
NMM does not support Microsoft Exchange Server 2000 or earlier.
that are supported by the NMM client.
NMM and Exchange Server support for L18N
Microsoft requires Exchange Server 2003 Storage Group names to be in the same
locale as the Microsoft Exchange Server.
Exchange Server backup and recovery types
NMM supports several types of backup and recovery:
194
◆
Backup Level
Full and incremental for Microsoft Exchange Server 2007. Full only for Microsoft
Exchange Server 2003.
◆
Microsoft Exchange Server 2003:
• VSS Shadow Copy Service (VSS)
Uses Microsoft-supplied Exchange VSS writers to back up and restore
Exchange databases and transaction log files.
• Point-in-time snapshot
A backup of exact copies of files and open files.
• Exchange Server backup
Backup of Exchange Server databases, logs, and storage groups.
• Storage group backup
Backup of an individual storage group by creating a save set that specifies just
that storage group.
• Roll-forward recovery
This recovery can be performed when the current log LUN is available. No
data is lost when restoring from backup. Recovers the Exchange Server data
from the last backup, and allows granular recovery selection down to the
individual mailbox database level. Deleted email is not recovered.
• Point-in-time recovery
Use when log files have been lost. All new data after the last backup is lost only the transaction files that were part of the backup set are restored. Deleted
mail can be recovered.
• Rollback restore
Also known as a “destructive recovery,” because all changes that were made
after the selected snapshot was taken are overwritten. The whole snapshot
must be selected for recovery; storage groups within a snapshot cannot be
selected for recovery. A single storage group can be recovered if the save set
specified only that storage group for backup.
◆
NMM supports the backup and recovery options listed for Exchange Server 2003,
and in addition supports:
• Backup in a Cluster Continuous Replication (CCR) environment that is
running on the active node or the passive node.
• Backup in a Local Continuous Replication (LCR) environment of the
production data, but not of the replicated data.
• Database recoveries to the Exchange Recovery Storage Group (RSG).
• Exchange RSG mailbox browsing; mailbox, folder, and message recovery.
• Exchange database recoveries to alternate servers and alternate storage
groups.
Updating and modifying Exchange credentials
Exchange credentials are configured during NMM installation. Usually, you will have
a specific account that you run your Exchange Services under. If Exchange Server is
195
installed after NMM is installed, or if the Exchange credentials have changed, NMM
must be updated with the new credential information.
Modifying Exchange Server 2003 credentials
When installing NMM or when running the nwexinfo utility, enter the correct
credentials. Otherwise backup of Client resource fails, even when the backup does
not include Exchange data.
To update Exchange Server 2003 credentials, run the nwexinfo.exe utility.
The EMC NetWorker Module for Microsoft Applications Release 2.2 Installation Guide
provides more information about initial setup of Exchange credentials during NMM
installation.
Adding Exchange Server 2007 credentials
If Exchange Server 2007 is installed after NMM is installed, then NMM must be
reinstalled after Exchange is installed. During NMM installation, add Exchange
credentials.
provides more information about initial setup of Exchange credentials during NMM
installation.
Modifying Exchange Server 2007 credentials
NMM installation configures your Exchange Server access with the username and
account information you specify. If the credentials for the Exchange Server 2007
account change after the initial NMM installation, then NMM must be updated with
the new credentials. You can use the following procedure to change the user account
or password information for the COM+ component and the Replication Manager
Exchange Interface Service.
To update Exchange Server 2007 credentials:
1. Start component services.
2. Expand Computers > My Computer > DCOM Configuration.
3. Right-click Replication Manager Exchange Interface and select Properties from
the context menu.
4. Click the Identity tab.
5. Modify the domain user account or password and click OK.
This will also modify the account information for the Replication Manager Exchange
Interface Service.
Compatibility with previous Exchange Server backups
The Exchange database and logs that are backed up by using previous releases of
NetWorker software may not be restored by using NetWorker Module for Microsoft
Applications. After upgrading to NMM, perform a full backup of Exchange Server in
order to use NMM to perform point-in-time restores in the future:
196
◆
To restore Exchange backups created with NetWorker Module for Exchange 5.x,
use NetWorker Module for Exchange. After the recovery, perform a full backup of
the Exchange server with NMM.
◆
To restore backups created with NetWorker VSS Client for Microsoft Windows
1.0, including backups created before the upgrade to NetWorker Module for
Microsoft Applications 2.0, use NMM.
Note: Upgrades to Microsoft Exchange Server, including Exchange Service Pack upgrades, do
not allow recovery from a backup created before the upgrade. After an Exchange Server
upgrade, perform a full backup to ensure that you have a new recoverable backup.
Exchange Server data backed up by NMM
NetWorker Module for Microsoft Applications supports full and incremental
backups of Microsoft Exchange Server 2007, and full backups of Microsoft Exchange
Server 2003. It backs up the following Exchange Server data:
◆
Mounted Exchange Server databases
◆
Transaction log files
◆
Checkpoint files in a storage group
Components used by NMM for Exchange Server backup and recovery
NMM uses the following commands and Exchange Server -specific attributes to back
up Exchange data:
◆
Application writers
NMM uses the application writers provided by Microsoft Exchange Server. There
is one writer for Exchange 2003, and two for Exchange 2007: an Information Store
writer and a Replication Service writer. The Replication Service writer is only
used in a CCR environment.
◆
Nsrsnap_vss_save.exe
NMM uses this command to initiate the VSS-based backup. This command can be
used with options for database consistency and Jet error checking.
◆
NSR_CHECK_JET_ERRORS
Option used with nsrsnap_vss_save.exe to detect Jet errors in the Exchange
backup. NSR_CHECK_JET_ERRORS is run before committing the shadow copy
for Exchange.
◆
eseutil
A program provided by Microsoft Exchange server to check that the database and
log files do not contain errors, and that the Exchange server is valid before
performing the restore.
◆
Nsrsnap_vss_recover.exe
NMM uses this command to initiate the VSS-based recovery. This command
cannot be run manually by users.
◆
NSR_MOUNT_DB
Specifies whether the database should be online or offline after restore.
197
◆
NSR_MOVE_TRANSLOGS
Specifies the type of Exchange recovery, either replay logs or replace logs for
point-in-time recovery.
◆
NSR_VIRT_SERV
Specifies the name of the Exchange Virtual Server Client resource in a CCR
passive node backup.
◆
NSR_INDEX_CLIENT
Specifies that the backup is to be indexed under the client name of the Exchange
Virtual Server, not the passive node client name in a CCR passive node backup.
Specifying Exchange save sets for application data
Table 35 on page 198 lists the Exchange save set syntax to specify for supported types
of Exchange data. Specify Exchange data save sets in the Save Set attribute of a Client
resource. save set syntax.
Table 35
Exchange save set syntax
Save set syntax
Exchange Server 2003,
all Exchange storage groups.
APPLICATIONS:\Microsoft Exchange Writer
Exchange Server 2003,
a specific Exchange storage group.
For example, a storage group named
first_group.
APPLICATIONS:\Microsoft Exchange
Writer\first_group
Some special characters in an Exchange storage group name
must be replaced with URL-encoded values. “URL encoding for
Exchange save sets” on page 199 provides more information
about URL encoding.
Exchange Server 2007 (Information Store),
all Exchange storage groups.
APPLICATIONS:\Microsoft Exchange 2007
Exchange Server 2007 (Information Store),
a specific storage group.
For example, a storage group named
first_storage_group.
APPLICATIONS:\Microsoft Exchange
2007\first_storage_group
Some special characters in an Exchange storage group name
must be replaced with URL-encoded values. “URL encoding for
Exchange save sets” on page 199 provides more information
about URL encoding.
Displaying valid Exchange data save sets
When the NMM client is installed on the Exchange server, you can display a list of
Exchange save sets available for backup.
To display a list of the Exchange save sets that are available for backup:
1. Open a command prompt on the Microsoft Exchange server.
application data save set names:
• If the Microsoft Exchange server is a standalone node or LCR active node,
type:
nsrsnap_vss_save -?
198
• If the Microsoft Exchange server is on an Exchange 2003 Cluster node, or
Exchange 2007 SCC node, type:
nsrsnap_vss_save -c virtual_server_name -?
• In an Exchange 2007 CCR environment, type the following on both the active
and passive nodes as one line:
nsrsnap_vss_save -c physical_server_name -A
NSR_VIRT_SERV=virtual_server_name -?
3. Press Enter.
"APPLICATIONS:\Microsoft
Exchange
Exchange
Exchange
Exchange
Writer"
Writer\first_group"
2007"
2007\second_group"
must be typed on a separate line. In the Save Set attribute you can specify either
the writer name or the storage group names, but not both in the same save set.
URL encoding for Exchange save sets
characters when naming Exchange storage groups.
For example, a storage group named:
Accounting\Bellevue
APPLICATIONS:\Microsoft Exchange 2007\Accounting%5CBellevue
Table 36 on page 199 lists the most commonly used special characters and their
URL-encoded values.
Table 36
Special character
URL-encoded value
Special character
URL-encoded value
\
%5C
?
%3F
/
%2F
]
%5D
"
%22
[
%5B
%
%25
}
%7D
#
%23
{
%7B
&
%26
^
%5E
<
%3C
‘
%60
>
%3E
|
%7C
199
Exchange application information variable settings
Table 37
200
Exchange application information variable settings (page 1 of 2)
Attribute name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
application information attribute client that moves snapshots from
• The local host
client.
If you are setting up an Exchange Client
a proxy client. A proxy client requires
hardware for transportable snapshots. If
a proxy client is specified, Exchange
proxy client. Related messages are
written to the Replication Manager log
files (erm_clientxx.log) on the proxy
client.
Exchange Client resource, consistency
checks are performed on the Exchange
server. Related messages are logged on
the Client resource host in both the
nmm.raw file.
NSR_DM_PORT=value
Specifies the control port number to
use in communicating with the data
mover client. This is the control port
that the Replication Manager (RM)
client service is running on.
EMC NetWorker Module for
Microsoft Applications Release 2.2
Installation Guide provides more
information about RM ports.
The RM client control port number. For
example:
NSR_DM_PORT=6728
An RM client control port number must
be specified, and must match the value
specified during installation. The default
value that NMM uses during installation
is 6728. If you specify a different value
during installation, you must specify that
value here.
NSR_ALT_PATH=value
Specifies the path on which to
mount Exchange snapshots during
the backup process. Exchange
snapshots are mounted so that the
database consistency check utility,
eseutil.exe, can be run.
A mount path.For example:
NSR_ALT_PATH=G:\mount_replica\
If a proxy client is specified in the
NSR_DATA_MOVER attribute, enter the
mount path on the proxy client.
Table 37
Exchange application information variable settings (page 2 of 2)
Attribute name
Description
Values
NSR_CHECK_JET_ERRORS
=value
Specifies which JET database errors
to check for in the event log before
taking a snapshot.
• 447 (Exchange Server 2007 only)
• 448 (Exchange Server 2007 only)
• 1018
• 1019
• 1022
• all
• none
Values can be separated with a comma,
for example:
NSR_CHECK_JET_ERRORS=1018,
1019
The default is all.
Note: If errors are found, the
snapshot will fail.
NSR_VIRT_SERV=value
Specifies the name of the Exchange
Virtual Server Client resource in a
CCR passive node backup.
Note: Do not use this attribute for
CCR active node backup.
NSR_INDEX_CLIENT=value
Specifies to index the backup under
the client name of the Exchange
Virtual Server, not the passive node
client name in a CCR passive node
backup.
Note: Do not use this attribute for
CCR active node backup.
The name of the Exchange Virtual
Server. For example:
NSR_VIRT_SERV=virtual_server_
name
This value is required in a CCR passive
node backup.
The name of the Exchange Virtual
Server. For example:
NSR_INDEX_CLIENT=virtual_
server_name
This value is required in a CCR passive
node backup.
Serverless backup
serverless backups. In a typical Exchange setup with production hosts, serverless
backups free the application server from much of the processing involved in a
nonpersistent or instant snapshot backup. In a serverless backup, a snapshot is
created on the application server and then immediately mounted on the proxy client
for further processing.
Serverless backups are particularly useful when there is additional processing
involved in a snapshot backup. For example, to help determine whether a snapshot of
a Microsoft Exchange database is consistent, the Exchange utility, eseutil, should be
run against the snapshot. Running eseutil can be disk and CPU intensive. Offloading
the work from the Exchange server to a proxy client frees resources on the Exchange
server.
A serverless backup can be rolled over to a conventional backup and the backup can
remain on the mounted storage disk volume as a persistent snapshot.
201
Snapshot process in an Exchange Server production and proxy client setup
Figure 32 on page 203 illustrates a LAN-based Exchange Server configuration with a
storage area network and a proxy client. Both the Exchange server and the proxy
client have the NetWorker Module for Microsoft Applications client installed. The
proxy client frees resources on the Exchange server by offloading from the Exchange
Server the work of processing and backing up snapshots. When a backup operation
uses a proxy client, it is known as a serverless backup.
1. The NetWorker server initiates the process by contacting the Exchange server
2. The Exchange server with the NMM client software creates a snapshot of the data
on the storage volume.
4. NMM runs the eseutil consistency check for Exchange on the proxy host.
5. If the consistency check is successful, NMM commits the shadow copy.
6. NMM instructs Exchange to truncate the log files.
7. The proxy client uses a snapshot in primary storage to transfer the data over the
LAN and into a conventional backup medium such as a file type disk, advanced
file type disk, or tape.
The EMC VSS provider is used for the transportable snapshot.
When the administrator moves the shadow copy to the traditional storage medium
connected to the proxy host, NMM provides the option to move the transportable
snapshot from the proxy host to the traditional storage medium. For this setup, the
administrator must configure a SAN storage node as a proxy client.
202
NetWorker Server
7.3.3 or later
LAN
Exchange Server
Proxy client
SAN
Primary storage
(For example,
S4
S5
S1
S3
NetWorker
Storage
Node 7.3.x
Traditional
storage
S2
GEN-000749
Figure 32
Snapshot process with Exchange production server and proxy client
Best practices for Exchange backup and recovery
In addition to “Best practices and considerations for backing up application data” on
page 50, this section lists the best practices and recommendations to follow when
using NetWorker Module for Microsoft Applications to back up and recover
Exchange Server.
Backups should be grouped with no more than 10 Storage Groups at a time
When backing up Exchange 2007, backups should be grouped with no more than 10
Storage Groups at a time. This is especially important when parallel consistency
checking has been turned on. Grouping more than 10 Storage Groups consumes the
available resources of the CPU, I/O and memory, which can lead to backup failures.
Use consistency check parameters to set threading and throttling to handle performance issues
Exchange backup performance is greatly affected by the consistency check utility. The
larger the databases and logs, the longer it takes to run the consistency checker. NMM
2.2 provides the options to run the consistency check on multiple storage groups in
parallel. NMM provides the ability to set threading and throttling to handle
performance issues through the use of the -A attribute values for eseutil
configuration settings.
Note: If multiple consistency checks are run in parallel without setting proper throttle limits, it
may cause performance problems on the proxy system. In extreme cases, this can cause I/O
bottlenecks so severe that consistency checks will fail and other operations with the system are
adversely affected.
203
The following attributes have been added. These are set in the Application
Information attribute of the Client resource:
◆
NSR_ESE_UTIL_SEQUENTIAL — Set to true or false to specify if eseutil should
be run sequentially (single threaded) or in parallel against multiple SG's
(multi-threaded).
◆
NSR_ESE_UTIL_THROTTLE — Set to true or false to specify if eseutil should be
throttled.
◆
NSR_ESE_THROTTLE_IOS — Number of I/O's between pauses when
throttling, value range of 100 < 10000. the default value is 100.
◆
NSR_ESE_THROTTLE_DURATION — Duration of pause in milliseconds when
throttling, value range of 1000 < 60000. The default value is 1000.
For example:
-A NSR_ESE_UTIL_SEQUENTIAL=false -A NSR_ESE_UTIL_THROTTLE=true
-A NSR_ESE_THROTTLE_IOS=500 -A NSR_ESE_THROTTLE_DURATION=5000
Additional Exchange backup and recovery best practices
◆
Backup time can be reduced by spreading data across different storage groups.
For example, it takes longer to back up a single Storage Group with 200 GB of
data, than it does back up 200 GB of data spread across 10 different Storage
Groups.
◆
Avoid using mount points for creating Storage Groups. It takes longer to run a
snapshot operation for a Storage Group with mount points, than it does for a
Storage Group with no mount points.
◆
Avoid using copy-on-write for Microsoft Exchange snapshots for anything other
than temporary snapshots. Copy-on-write can be used if data is to be backed up
with the “always” setting in the Snapshot Policy in NetWorker Management
Console. Otherwise, for reliable snapshots, split-technology such as cloning or
mirroring is required.
◆
In addition to your scheduled full backups, you should also perform a full
backup copy of Exchange Server after:
• Every successful recovery.
• Upgrading to NMM from previous releases of NetWorker clients.
• The database directory path is changed or the storage group Log Path or
System Path are changed.
Note: If a full backup is not performed after the directory path or storage group Log
Path or System Path are changed, and a recovery is necessary, NMM will attempt to
restore to the old location and recovery will fail. To work around this, perform a
recovery to an alternate location as described in “Performing a directed recovery” on
page 238.
204
◆
Ensure that all databases in a specified storage group are mounted before backing
up the Exchange servers. Unmounted databases are not backed up.
◆
Ensure that database (mailbox and public folder) files and transaction log files
reside on separate volumes for backup, otherwise the backup will fail.
◆
If Exchange objects like storage groups and databases are deleted by using
Exchange System Manager in Exchange Server 2003, or Exchange Management
Console in Exchange Server 2007, these objects cannot be recovered until disaster
recovery is performed. Objects from the Exchange Server should not be deleted
unless they no longer need to be recovered.
◆
After upgrading to the NMM client, you cannot recover Exchange backups that
were performed with a previous version of the NetWorker client. To ensure that
all Exchange data can be recovered to the point-in-time of the upgrade, perform a
full backup of Exchange data immediately after upgrading to the NMM client.
◆
In Exchange Server 2003 backups, save sets and backup groups that include the
Exchange writer cannot include other application writers or the SYSTEM
COMPONENTS writer. The Exchange writer can however, be included in the
same save set with volume components such as D:\.
◆
In Exchange Server 2007, save sets and backup groups that include the Exchange
writer cannot include any other volumes, applications, or non-Exchange items in
the save set.
◆
The Exchange consistency checker utility (eseutil.exe) must be installed on the
proxy host. Ensure that the version of eseutil (Service Pack level) is the same as
the Exchange server. For example, if Exchange Server 2003 SP1 is installed on the
Exchange server, then the version of eseutil that is installed on the proxy client,
the Data Mover host, must be from the Exchange Server 2003 SP1 management
utilities. Failure to do so causes the consistency checker utility to report errors
even when the database is valid.
◆
While performing a rollback restore of the Exchange server, all applications that
are used by Exchange should be closed. Also, ensure that the volume that
contains the Exchange data is not in use.
◆
If a rollback restore of a particular storage group is performed, ensure that this
volume is not used by any other storage group that is not part of the recovery.
205
Performing Exchange Server backups
NMM supports several types and levels of backup:
◆
Full and incremental backups. Incremental backups are available for Microsoft
Exchange Server 2007, but only full backups are available for Microsoft Exchange
Server 2003.
◆
◆
Backup of Exchange databases, logs, and storage groups.
◆
Backup in a Cluster Continuous Replication (CCR) environment on the active
node and the passive node.
◆
Backup in a Local Continuous Replication (LCR) environment of the production
data, but not of the replicated data.
Configuring Exchange Server backups
supported by NMM. Chapter 2, “Configuring a Scheduled Backup” describes these
tasks in detail:
◆
◆
NMM supports incremental backups through the snapshot policies.
◆
◆
The next task is specific to configuring an Exchange Client resource:
“Configuring an Exchange Client resource” on page 208
Note: Once you have started a snapshot of a save group, do not interrupt or halt the snapshot
process. For example, in Exchange backup, the nsrsnap_vss_save.exe process on the
production server and the eseutil process on the proxy may continue to run after the snapshot
is halted. Any attempt to stop a save group in NetWorker Management Console will take a
long time to complete.
Incremental backup of Exchange Server 2007 requires NetWorker 7.5.1, or NetWorker 7.4
The initial release of NetWorker 7.5 prevents Exchange incremental backups from
occurring. Only Exchange Full backups will be requested by the server. A fix was
provided for this in the NetWorker 7.5.1. There will be no NetWorker 7.5 hotfixes, so
in order to perform Exchange Server incremental backups, NetWorker 7.5 users must
obtain and install the NetWorker 7.5.1.
NetWorker 7.4 users are not impacted, and no upgrades or hotfixes are required in
order to perform Exchange Server incremental backups.
Default backup schedule now includes incremental backups in NMM 2.2
NMM 2.2 supports incremental backups of Exchange Server 2007. Because NMM 2.1
and earlier releases did not support incremental or differential backups, the Default
NMM backup schedule was Full only. When you upgrade the client machine to
NMM 2.2 from an earlier version, the Default schedule is automatically changed to
206
perform a full backup on Sundays and incremental backups the rest of the week. If
you used a Default schedule, but the schedules have not been set for rollovers, those
save sets are automatically promoted to full backups.
If you upgraded the client to NMM 2.2 and use a Default schedule, but do not want to
use or include incremental backups, configure your Default backup schedule to take
only Full backups.
Configuring incremental backups
Incremental backups improve backup performance by reducing the amount of data
that is processed. However, recovery may take longer because you are recovering the
last full backup plus each incremental backup that was performed since the full
backup.
For Exchange incremental backups, the snapshot should be rolled over to the
NetWorker server. If using a standard snapshot policy, specify either Serverless
Backup or Rollover. If using a custom snapshot policy, specify Backup
Snapshots=All.
Performing Exchange Server backups
207
Configuring an Exchange Client resource
A Client resource specifies what to include in a snapshot of a NMM client. Client
policies. An Exchange Client resource can specify to perform a full backup, or back
up a storage group. “Configuring an Exchange Client resource” on page 208 provides
the steps for configuring standard Exchange Client resources for standalone, LCR, or
single copy cluster (SCC).
“Configuring backup for an Exchange CCR active node” on page 211 provides the
steps for configuring an Exchange Client resource for an Exchange CCR. The steps
vary depending on whether the node is the active node or passive node, whether a
data mover is used, or you want to use deduplication.
An Exchange Client resource can also be configured to use deduplication. Chapter 3,
“Deduplication Backup and Recovery,” provides in-depth information about
deduplication. “Configuring deduplication backup for a passive node in a CCR
environment” on page 218 provides the additional steps for creating an Exchange
Client resource in a CCR environment.
To create an Exchange Client resource for Exchange standalone, LCR, or SCC:
Configuration.
4. In the Name attribute, type the fully qualified domain name (FQDN) of the
NetWorker client computer.
• To back up all Exchange Server 2003 data, specify the save set name:
APPLICATIONS:\Microsoft Exchange Writer
• To back up all Exchange Server 2007 data, specify the save set name:
• To back up a specific Exchange Server 2003 storage group, for example a
storage group named Group_1, specify the save set name:
APPLICATIONS:\Microsoft Exchange Writer\Group_1
208
• To back up a specific Exchange Server 2007 storage group, for example a
group named Group_2, specify:
APPLICATIONS:\Microsoft Exchange 2007\Group_2
• To back up all Exchange Server 2007 data in a CCR environment on the active
or passive CCR node, specify:
• To back up a specific Exchange Server 2007 storage group in a CCR
environment on the active or passive CCR node, for example a group named
Group_1, specify:
• To back up all Exchange Server 2007 data on a virtual server, on the active
CCR node, specify:
Note: When backing up an Exchange Server 2007 in a CCR environment on a passive
CCR system, you must use the NSR_INDEX_CLIENT attribute.
• To back up all Exchange Server 2007 data in a LCR environment, specify:
• To back up a specific Exchange Server 2007 storage group in a LCR
environment, for example a group named Group_1, specify:
The Exchange Server save set syntax is described in “Specifying Exchange save
Note: When backing up data for a Microsoft Exchange application, ensure that all
databases in a specified Storage Group are mounted. If at least one database in a selected
Storage Group is not mounted, the entire backup of the Storage Group will fail.
backslash (\), and a table of values is provided in “URL encoding for Exchange
save sets” on page 199.
will be added.
Note: The backup group must be a snapshot group.
209
type:
NSR_SNAP_TYPE=vss
Type Exchange application variable settings and their values on separate lines.
18. Verify that the Aliases attribute displays the NETBIOS name for the client. This
name should have already been filled in automatically.
19. Click OK.
Configuring backup for Microsoft Exchange Server 2007 CCR
The steps for configuring backup for Exchange CCR are the same as those for
configuring any Exchange Client resource, but with some additional settings for the
Application information.
Client resources required for active and passive nodes
In all Exchange cluster backups, whether for a passive node or active node, a client
resource is required to back up the Exchange server, and a client resource is required
for each physical machine. In an active node backup, the client resource is created for
the Exchange virtual server, so only one client resource is needed. In a passive node
backup, two client resources are required, one for the virtual server and one for the
physical machine. In the passive node backup, the virtual server client resource does
not need to be backed up by NetWorker, but is required so that the necessary index
entries are created within the NetWorker server.
The following procedures describe these different types of CCR Exchange
configurations:
210
◆
“Configuring backup for an Exchange CCR active node” on page 211
◆
“Configuring backup for an Exchange CCR passive node” on page 212
◆
“Configuring backup for an Exchange CCR active node with a data mover” on
page 214
◆
“Configuring backup for an Exchange CCR passive node with a data mover” on
page 215
◆
“Configuring deduplication backup for a passive node in a CCR environment” on
page 218
Configuring backup for an Exchange CCR active node
In an active node backup, only one client resource is required.
To create an Exchange Client resource if backup is performed from an Exchange CCR
active node:
Configuration.
4. In the Name attribute, type the <Exchange Virtual Server hostname>
9. In the Save Set attribute, specify APPLICATIONS:\Microsoft Exchange 2007.
Place multiple entries on separate lines:
will be added.
15. In the Backup command attribute, type the backup command:.
211
16. In the Application information attribute, type the following variables and
values:
NSR_SNAP_TYPE=vss
NSR_ALT_PATH=<mount_path_on_local_host>
Note: NMM client user interface can use either the NETBIOS name or the fully-qualified
domain name (FQDN) when connecting to the NetWorker server to browse backups.
19. Click OK.
Configuring backup for an Exchange CCR passive node
When backing up Exchange from the passive node in the Exchange CCR
environment, two client resources need to be created. The configuration of the two
client resources is similar, but there are differences depending on whether it is the
virtual server or physical node:
◆
A client resource using the passive node’s physical name in the Name attribute.
This backs up the Exchange database files, and must be backed up by NetWorker.
◆
A client resource using the Exchange virtual server name in the Name attribute.
This resource does not need to be backed up by NetWorker, but is required so that
the necessary index entries are created within the NetWorker server.
passive node:
Note: “Configuring deduplication backup for a passive node in a CCR environment” on
page 218 provides additional steps for enabling deduplication.
Configuration.
4. In the Name attribute, specify the physical node name for the passive node client
resource, or the Exchange server name for the virtual server client resource.
212
will be added.
values:
NSR_SNAP_TYPE=vss
NSR_ALT_PATH=<mount_path_on_local_host>
For a passive node client resource, but not the virtual server resource, also specify
the following:
NSR_VIRT_SERV=<Exchange_Virtual_Server_hostname>
NSR_INDEX_CLIENT=<Exchange_Virtual_Server_hostname>
17. Click the Globals (1 of 2) tab, and verify that the Aliases attribute displays the
NETBIOS name for the client. This name should have already been filled in
automatically.
18. Click OK.
213
Configuring backup for an Exchange CCR active node with a data mover
In an active node backup, only one client resource is required.
active node with a data mover:
Configuration.
4. In the Name attribute, type the <Exchange_Virtual_Server_hostname>
will be added.
214
values:
NSR_SNAP_TYPE=vss
NSR_ALT_PATH=<mount_path_on_proxy_client>
NSR_DATA_MOVER=<proxy_client_hostname>
NSR_DM_PORT=6728
Note: The default value in NMM installation is 6728. If you changed the DM Port number
to a different value in your NMM client configuration or installation, use that value for
NSR_DM_PORT.
19. Click OK.
Configuring backup for an Exchange CCR passive node with a data mover
When backing up Exchange from the passive node in the Exchange CCR
environment, two client resources need to be created:
◆
A client resource using the passive node’s physical name in the Name attribute.
This backs up the Exchange database files, and must be backed up by NetWorker.
◆
A client resource using the Exchange virtual server name in the Name attribute.
This resource does not need to be backed up by NetWorker, but is required so that
the necessary index entries are created within the NetWorker server.
passive node with a data mover:
Configuration.
4. In the Name attribute, specify the physical node name for the passive node client
resource. or the Exchange server name for the virtual server client resource.
215
will be added.
values:
NSR_SNAP_TYPE=vss
NSR_ALT_PATH=<mount_path_on_proxy_client>
NSR_DATA_MOVER=<proxy_client_hostname>
NSR_DM_PORT=6728
Note: The default value in NMM installation is 6728. If you changed the DM Port number
to a different value in your NMM client configuration or installation, use that value for
NSR_DM_PORT.
In addition, for the passive node only, specify:
NSR_VIRT_SERV=<Exchange_Virtual_Server_hostname>
NSR_INDEX_CLIENT=<Exchange_Virtual_Server_hostname>
216
19. Click OK.
217
Configuring deduplication backup for a passive node in a CCR environment
To perform deduplication backup of a passive node in a CCR environment:
1. In the server’s Administration interface, click the Configuration button.
2. Select Clients in the navigation tree.
The Clients table appears. It lists any de-duplication clients that have already
been created.
3. Right-click Clients in the navigation tree, or right-click any client in the Clients
table, and select New.
The General tab appears in the Create Client window.
4. On the General tab, enter the information as for any client, with the following
exceptions:
a. Clear the Scheduled backup checkbox.
b. In the Save set list, add the same save set list that is in the passive node.
5. Click the Apps & Modules tab:
a. In the Access area, leave the Remote user and Password fields empty.
b. In the De-duplication area, select the De-duplication backup attribute to
enable this client for deduplication backups.
c. Select the name of the deduplication node to which this client’s backup data
will be sent. This should be the same deduplication node specified for the
passive node.
This step links this client with its own deduplication node. Do not select the
name of a replication node here.
!
IMPORTANT
If the deduplication node for this client’s backup data changes, the next backup
performed must be a level 0 (full) backup.
6. Complete the remaining configuration information as for any regular client.
7. Click OK.
8. Select Groups in the navigation tree.
9. Add the new client resource to the group that contains the passive node backup.
218
Performing Exchange Server recovery
After the files are backed up, administrators can restore:
◆
All databases in a storage group.
◆
Individual mailbox or public folder databases.
◆
User email messages.
Several types of recovery are available:
◆
Point-in-time
◆
Rollback
◆
Roll forward
◆
Recovery to Recovery Storage Group (RSG) from snapshots or conventional save
sets
◆
Recovery to an alternate storage group
◆
Restore email messages from RSG databases
◆
Disaster recovery
Exchange recovery requirements and limitations
The following requirements must be met before recovering Microsoft Exchange data:
◆
The writer set or storage group save set must have been backed up:
APPLICATIONS:\Microsoft Exchange Writer save set (Exchange Server 2003
writer), APPLICATIONS:\Microsoft Exchange 2007 save set (Exchange Server
2007 writer), APPLICATIONS:\Microsoft Exchange Writer\first_group
(Exchange Server 2003 storage group), or APPLICATIONS:\Microsoft Exchange
2007\second_group (Exchange Server 2007 storage group). Chapter 2,
“Configuring a Scheduled Backup,” provides more information about backups.
◆
For each Exchange database to be recovered, the Exchange checkbox titled “This
database can be overwritten by a restore” must be selected in the Exchange
System Manager (Exchange Server 2003) or Exchange Management Console
(Exchange Server 2007) application. This is a Microsoft requirement. For
information about this checkbox, refer to the Microsoft Exchange documentation.
◆
When new Microsoft service packs for Exchange Server are installed, transactions
(log files) created with previous service pack versions of Exchange Server
generally cannot be recovered onto the updated Exchange Server. Microsoft
sometimes provides hotfix kits to address this Microsoft restriction.
Note: It is recommended that you perform a full backup after any Exchange server
software upgrade.
◆
If you are recovering Exchange data that was backed up from a legacy NetWorker
client by using the save set VSS USER DATA:\ in a NMM File System Recovery
Session, then the Exchange databases to which data is being recovered must be
unmounted before the recovery operation. After the recovery, the databases must
be mounted.
219
There are several limitations when backing up and restoring Exchange objects with
NMM:
◆
A single mailbox database cannot be restored by using point-in-time recovery of
Exchange, because it requires both logs and databases to be selected for restore.
VSS-marking semantics do not allow selecting logs for backup of restore. Logs are
included only when the a storage group is selected for either backup or restore.
Logs are not included when a database is selected.
◆
Roll-forward recovery may not be possible after point-in-time restore. After you
complete a successful point-in-time restore, perform a full backup of the
Exchange server so that you can perform roll-forward recovers.
◆
Microsoft Exchange server supports an RSG configuration where the RSG system
path restore location and RSG logs restore location can be different. NMM does
not currently support that configuration. Specify the same location for the RSG
system path and the RSG log path.
MAPI memory errors when performing item-level mailbox recoveries
There is a known memory error that may occur with Microsoft Exchange MAPI when
performing item-level mailbox recoveries. The error
MAPI_E_NOT_ENOUGH_MEMORY may be reported in the NMM logs.
When this error occurs, NMM recovers the mailbox item, but may lose properties
such as the original font and formatting.
This is a known problem, but Microsoft has no workaround or fix for this issue at this
time.
IPv6 issues affect browsing RSG mailboxes
There is a known issue with IPv6 affecting MAPI calls. NMM uses MAPI calls to
browse RSG mailboxes. This may occur when the NMM client is running Windows
Server 2008 x64 and Exchange Server 2007, and IPv6 is enabled. IPv6 is enabled by
default by Windows Server 2008 installation. When IPv6 is enabled on the client, and
there is no IPv6 network present, RSG mailbox browsing will fail. This can prevent
you from viewing and selecting RSG mailboxes. This problem must be fixed through
the registry, not the network connections dialog box.
To fix RSG browsing issue in the registry:
1. Open the registry.
2. Navigate to HKLM\System\CurrentControlSet\Services\Tcpip6\Parameters.
3. Edit or create the 32-bit DWORD value DisabledComponents, and enter the hex
value FFFFFFFF.
Quick recovery of a deleted mailbox
In the event that a user mailbox is deleted, you can quickly recover the mailbox
through the Exchange Management Console. This procedure is performed
independent of NMM backup or recovery operations.
To recover a deleted mailbox:
1. In Exchange Management Console, make sure that deleted user mailbox entry
appears under Disconnected mailbox under Recipient management.
2. Re-create the user mailbox with same name.
220
3. Disable the re-created user mailbox.
4. Open Disconnected mailbox under Recipient management; and then select the
user mailbox which was accidentally deleted.
5. Right-click the user mailbox and then select Connect.
6. Select User mailbox radio button, and then click Next.
7. Select Matching user, select User mailbox, and then click Next.
The deleted user mailbox is then re-created under Recipient management >
Mailbox.
Note: Recreated deleted mailboxes are not always immediately visible. It may take 15 minutes
to an hour before the mailbox is visible and accessible through Outlook.
Exchange recovery options
To specify recovery options for Exchange data:
3. From the left pane, select Exchange Recover Session, and then select one of the
following:
• Database Recover (default) to view Exchange database backups.
• RSG Mailbox Recover to browse and recover items from existing RSG
databases.
4. In the Exchange Server Session toolbar, click Recover Options.
5. Select the type of recovery to perform:
• To recover data for Exchange Server 2003 or 2007 and retain all existing
Exchange transactions that occurred after the time of backup, select Include
existing logs (Roll forward recovery).
“Roll-forward recovery” on page 222 describes this type of recovery in detail.
• To recover data for Exchange Server 2003 or 2007 up to the time of backup
only and to not include subsequent transactions, select Include logs only from
this restore (Point-in-time recovery).
If you select point-in-time recovery, you cannot select an individual database
for recovery. However, you can select to recover items at the Storage Group
level or later.
“Point-in-time recovery” on page 223 describes this type of recovery in detail.
6. To make the Exchange database available after the recovery, select Put databases
online after restore.
7. Click OK to close the Exchange Recover Options dialog box.
8. From the navigation tree, expand the Applications folder and the appropriate
Microsoft Exchange writer folder:
• Microsoft Exchange Writer for Microsoft Exchange Server 2003
221
• Microsoft Exchange 2007 for Microsoft Exchange Server 2007
9. Select the Exchange storage groups and databases that are to be recovered.
10. From the Exchange Recover toolbar, click Start Restore.
Roll-forward recovery
This option recovers the Exchange databases from the last full backup, and logs from
incremental backups. All existing Exchange transactions that occurred after the time
of the last backup are retained. This recovery is carried out when the Exchange server
or database fails but the current log LUN is available.
NMM restores the database files and transaction logs from backups and uses the
current logs on the server to roll the database forward. No data is lost by restoring
from backup. A single mailbox database can be marked for this type of recovery in
NetWorker Module for Microsoft Applications. While performing the roll-forward
recover, all of the mailboxes pertaining to that storage group are dismounted and
once the recovery is successful, all of the mailboxes are remounted.
Note: If you perform a roll-forward recovery, perform a full backup after the recovery. Doing so
enables you to recover data to a point-in-time that is after the point-in-time of the roll-forward
recovery, if necessary. Also, a roll-forward recovery is not possible after performing a
point-in-time recovery.
To perform a roll-forward recovery of an Exchange Server 2007 in a CCR
environment:
1. Perform the recovery from the active node, failing over the passive node to the
active node if necessary.
2. In the Exchange Management Console application, select the Microsoft
Exchange checkbox This database can be overwritten by a restore.
3. In the Exchange Management shell, stop replication to the passive node with the
suspend-StorageGroupCopy command.
For example:
Suspend-StorageGroupCopy -Identity "bv-hood-cluse12\First
Storage Group"
4. Open the NMM client.
5. In the main toolbar, click the Client list, and then select the client that is the
Exchange Virtual Server.
6. From the left pane, select Exchange Recover, and then select one of the following:
• RSG Recover to browse and recover items form existing RSG databases.
7. In Transaction Log File Replay, make sure Include existing logs (Roll-forward
recovery) is selected. This is the default setting, but it may have been changed if
the previous recovery was a point-in-time recovery.
8. From the navigation tree, expand the Applications folder and select the
Microsoft Exchange 2007 folder.
9. Select the Exchange Server items you want to restore.
222
10. From the Exchange Recover Session toolbar, click Recover.
The Exchange Recovery Options Summary dialog box is displayed.
11. Review the options:
• If you need to change the options, click Recover Options. This opens the
NetWorker Recovery Options dialog box, where you can change settings on
the General, NetWorker, Security, and Exchange tabs.
• If the options are okay, click the Start Recover button. This closes the
Exchange Recovery Options dialog box, and starts the recovery.
12. Restore replication between the nodes with the resume-StorageGroupCopy
command.
For example:
resume-StorageGroupCopy -Identity
"bv-hood-cluse12\First Storage Group”
Point-in-time recovery
This recovery type is useful when log files have been lost or databases must be
restored to a specific backup time. At this level, data can be selected at a Storage
Group level, but not at an individual database level. Point-in-time recovers Exchange
data only up to the time of the backup that is being recovered, and does not include
subsequent transactions. All new data after that backup is lost.
When point-in-time recovery is used, only the transaction log files that were part of
the backup set are restored. Additional log files generated since the time of backup
are not restored, and the databases are recovered only to the point of the backup.
With this type of recovery the administrator gets back the deleted mail because the
transaction logs are restored.
In order to perform point-in-time restore, the administrator must first select the
whole storage group and then perform the recover. The log files currently on disk are
moved to another directory before the restore takes place. They can be manually
deleted by the user if they are no longer needed after a successful restore.
To perform a point-in-time recovery of an Exchange Server 2007 in a CCR
environment:
1. Perform the recovery from the active node, failing over the passive node to the
active node if necessary.
2. In the Exchange Management Console application, select the Microsoft
Exchange checkbox This database can be overwritten by a restore.
3. In the Exchange command shell, stop replication to the passive node with the
suspend-StorageGroupCopy command.
For example:
suspend-StorageGroupCopy -Identity "bv-hood-cluse12\First
Storage Group"
4. Open the NMM client.
5. In the main toolbar, click the Client list, and then select the client that is the
Exchange Virtual Server.
223
6. From the left pane, select Exchange Recover, and then select one of the following:
• RSG Recover to browse and recover items form existing RSG databases.
7. In Transaction Log File Replay, select Include only logs from this restore
(Point-in-time recovery).
8. From the navigation tree, expand the Applications folder and select the
Microsoft Exchange 2007 folder.
9. Select the Exchange Server items you want to restore.
10. From the Exchange Recover Session toolbar, click Start Restore.
11. After restore, on the passive node manually delete log files and database file.
12. On the passive node, in the Exchange command shell, re-seed the passive node
with the Update-StorageGroupCopy command.
For example:
Update-StorageGroupCopy -Identity
"bv-hood-cluse12\First Storage Group"
13. Restore replication between the nodes with the resume-StorageGroupCopy
command.
For example:
resume-StorageGroupCopy -Identity
"bv-hood-cluse12\First Storage Group”
Mounting the database after recovery
Certain pre- and post-restore operations provided for Exchange enable quick access
to email after recovery. Until Exchange databases are remounted, email is not
available for users to browse and verify. NMM provides options to automatically
mount Exchange databases after restore. The Exchange administrator can decide if
restore of a particular Exchange database is required and if automatic mounting of
the database must be done after the restore is complete. If this option is specified,
then the database is mounted after successful restore of the Exchange Server. If this
option is disabled, then the administrator must manually mount the database. By
default, this option is enabled in NMM.
Select or disable this option with the Put databases online after restore checkbox as
described in “Exchange recovery options” on page 221.
Rollback restore of Exchange data by using NMM
This is a volume-level operation so any other data on the volume will also be lost
during a rollback.This recovery is known as destructive recovery because all changes
that were made after the selected snapshot was taken are overwritten.
Note: NMM does not support rollback restore for backups performed with
EqualLogic or Celerra hardware. Only conventional restore and snapshot restore are
supported for those hardware.
Rollback recovery of Exchange does not allow granular selection of storage groups
for recovery, the whole snapshot must be selected for recovery.
224
To perform rollback restore of a particular storage group, the storage group must
have previously been backed up by itself, and by using a save set that specified that
particular storage group. For example, to perform a rollback restore of the storage
group SG1, the storage group would have had to been previously backed up with the
save set "APPLICATIONS:\Microsoft Exchange Writer\SG1" (Exchange Server 2003)
or "APPLICATIONS:\Microsoft Exchange 2007\SG1" (Exchange Server 2007).
Exchange rollback operation requirements
To perform an Exchange rollback operation the following requirements apply:
◆
The snapshot should be a transportable snapshot.
◆
The rollback option is not available if the snapshot is taken by the software
provider.
◆
A storage array descriptor file, sa.ini, must be configured with details of the type
of storage used. “Storage array descriptor file” on page 327 provides detailed
information about creating and using this file.
◆
The Microsoft Exchange checkbox titled “This database can be overwritten by a
restore” must be selected for each Exchange database that is to be recovered. This
checkbox can be found by using the Exchange SYSTEM Manager application in
Exchange Server 2003 or Exchange Management Console in Exchange Server
2007.
◆
The administrator must manually mount the Exchange databases after recovery.
Automatic mounting of database is not available for rollback restore.
◆
“Roll back a snapshot” on page 327 provides more information about rollback
recovery, creating a storage array descriptor file, and the steps to perform a
rollback recovery.
To perform a rollback recovery of Exchange Server 2003 or 2007, follow the steps in
“Performing a rollback recovery” on page 329.
To perform a rollback recovery of an Exchange Server 2007 in a CCR environment:
1. Make sure that the current Exchange active node is the node where the files
existed at save time.
Failover the virtual Exchange server to the passive node where the backup was
taken, if necessary.
2. Follow the steps in “Performing a rollback recovery” on page 329 by using the
following settings:
• The Exchange Server 2007 recovery type is Point-in-time.
• Browse and select the items for recovery from the client type Exchange Virtual
Server.
225
Recovery to storage groups
NMM can recover an NMM point-in-time or rollover backup to an Exchange Server
2007 Recovery Storage Group (RSG), the original storage group, or to an alternate
storage group. When restoring from a point-in-time backup, files are copied from a
snapshot to the specified location. When restoring from a rollover, files are copied
from a NetWorker device such as tape, file, or adv_file, to the specified location.
!
IMPORTANT
When you use the restore to an RSG feature, always delete the existing RSG and
then create a new clean RSG.
Selecting a storage group restore destination
NMM provides several options for selecting the target storage group. These options
are specified in the Select Storage Group Restore Destination dialog box, which is
accessed in the Exchange Recover Session. Figure 33 on page 226 displays the options
available in this dialog box.
Figure 33
Select Storage Group Restore Destination dialog box
The steps to specify which location to restore to are described in the separate
procedures for each location type:
◆
“Recovery to the original storage group” on page 227
◆
“Recovery to an Exchange RSG” on page 228
◆
“Recovery to an alternate storage group” on page 235
The Select Storage Group Restore Destination dialog box provides the following
settings and information:
◆
226
<Storage Group Name> Storage Group (Source) — This box displays the name of
the storage group selected in the Browse tab in the Exchange Recover Session
window.
◆
Restore To (Destination) — Specifies where you want the storage group restored
to.
◆
Exchange Server — By default this is server where the source storage group is
located. You can select a different server here if you are recovering to an RSG or
alternate storage group that is not on the source server.
◆
Original Storage Group — Specifies to restore the content to the original storage
group, which will overwrite the content currently in that storage group.
◆
Recovery Storage Group — Specifies to restore the storage group to a separate
RSG you have already created. Recovering to an RSG allows you to browse and
select items to restore without overwriting the entire storage group. When this
option is selected, the Check Recovery Storage Group box is displayed. The
Check button is used to verify that the Recovery Storage Group exists and
databases are configured properly.
◆
Alternate Storage Group — Specifies to restore databases to a different storage
group. When this option is selected the Select Alternate Storage Group box
provides a list box where you can select from the available storage groups
available on the Exchange Server listed in the Restore To box.
Note: Exchange Server 2007 does not support restoring a public folder to the RSG. If a storage
group contains a public folder, it cannot be selected, only the mailboxes within that storage
group can be selected.
Recovery to the original storage group
Recovery to the original storage group overwrites all of the current content of the
selected databases in the storage group.
To recover to the original storage group:
3. From the left pane, select Exchange Recover, and then select Database Recover
4. In the navigation tree, expand the Exchange server you want to recover from.
NMM displays the contents of the storage groups that have been backed up.
5. Right-click the storage group that you want to recover, and then click Restore To.
6. In the Select Storage Group Restore Destination dialog box, verify that Original
Storage Group is selected and then click OK.
7. Click Start Recover.
8. After the restore, the administrator may need to mount the databases if the option
to put databases online was automatically selected.
227
Recovery to an Exchange RSG
NMM does not back up the contents of an RSG. The RSG is used only as a location for
restoring backups from normal storage groups.
There are two parts to the recovery process:
◆
“Creating an RSG and adding target databases” on page 228
◆
“Recovering to the RSG” on page 229
Creating an RSG and adding target databases
Before NMM can restore to an RSG, an RSG must be created and linked to a storage
group. Exchange Server 2007 only supports one RSG at a time per server.
!
IMPORTANT
then create a new clean RSG. If you are performing a directed recovery, delete the
existing RSG on the original server as well as the RSG on the destination server.
There are two methods available within Microsoft Exchange to configure an RSG
database:
◆
Exchange Management Shell at command line. Recommended.
◆
Exchange Management Console Database Recovery Management tool.
To configure an RSG with Exchange Management Shell:
1. On the Microsoft Exchange Server, select the Start menu and then select Exchange
Management Shell.
2. Use the new-storage group command to create the Recovery Storage Group:
new-storagegroup –server exch_server –name rsg_name
–LogFolderPath rsg_path –SystemFolderPath rsg_path –Recovery
where:
• exch_server is the Exchange 2007 server name.
• rsg_name is the desired RSG name.
• rsg_path is the path to the RSG directory.
The –Recovery option is also needed to specify that an RSG is created.
For example:
new-storagegroup –server bv-rmx64-1 –name SG2_RSG
–LogFolderPath T:\exchsrvr\SG2\SG2_RSG –SystemFolderPath
T:\exchsrvr\SG2\SG2_RSG –Recovery
3. Use the new-mailboxdatabase command to add target databases to the RSG:
new-mailboxdatabase -Name db_name -MailboxDatabaseToRecover
db_name -StorageGroup exch_server\rsg_name -EdbFilePath
edb_path
where:
• db_name is the name of the database that will be recovered.
• -EdbFilePath is the path of the database in the RSG, not the original storage
group.
228
• edb_path is the path to the database file in the RSG directory.
Use the same name for the RSG database as the original database to avoid
confusion.
For example:
new-mailboxdatabase -Name DB1 -MailboxDatabaseToRecover DB1
-StorageGroup bv-rmx64-1\SG2_RSG -EdbFilePath
U:\exchsrvr\SG2\SG2_RSG\DB1.edb
4. Repeat the new-mailboxdatabase for each database that will recovered. Be sure
to specify the correct database name and edb path.
5. NMM allows the administrator to leave the database offline or have it
automatically put online after the restore. If you choose to leave the database
offline after the restore it will need to be manually mounted. This can be done
with the mount-database command:
mount-database –Identity exch_server\sg_name\db_name
For example:
mount-database –Identity bv-rmx64-1\SG2_RSG\DB1
Databases can be dismounted with the dismount-database command:
dismount-database –Identity exch_server\sg_name\db_name
For example:
dismount-database –Identity bv-rmx64-1\SG2_RSG\DB1
6. To repeat an RSG restore by using the same database after it has been mounted,
set the database can be overwritten property:
set-mailboxdatabase –Identity “exch_server\rsg_name\db_name”
–AllowFileRestore:$true
For example:
set-mailboxdatabase –Identity “bv-rmx64-1\RSG\DB1”
Recovering to the RSG
When an RSG has been created and linked to a storage group, it is enabled in NMM
and can be selected as a restore destination. After databases have been recovered to
the RSG, NMM can perform item-level recovery for folders or individual items.
!
IMPORTANT
then create a new clean RSG.
To recover to an RSG:
1. Open the NetWorker User UI and select the NetWorker server on which the
3. From the left pane, select Recover Session > Exchange Recover, and then select
Database Recover. This displays the available Exchange database backups.
229
Figure 34 on page 230 shows the “SG2” storage group has been selected, so the
“First Storage Group” and “Public_SG” are unavailable and cannot be selected.
This prevents simultaneous restore of one storage group to the RSG and another
storage group to the original location.
Figure 34
Available and unavailable storage groups.
5. Right-click the storage group that is linked to the RSG.
Figure 35 on page 230 shows the context menu with the Restore To command
available.
Figure 35
230
Restore To command on context menu
6. Select Restore To, and then select RSG Configuration.
This verifies that all RSG databases exist and can be overwritten. It checks all
databases, in case the adminstrator marks the entire storage group. If an error
message like the one in Figure 36 on page 231 appears, there is a configuration
problem. The configuration problem must be fixed before the database with the
configuration problem can be restored. Then the restore operation can be started.
In the example shown in Figure 36 on page 231, DB1 and DB2 could be restored,
but DB3 cannot until the configuration problem is fixed.
Figure 36
RSG Configuration error message
When a message like the one in Figure 37 on page 231 appears, the RSG
configuration is correct.
Figure 37
RSG Configuration correct message
RSG configuration is also checked when a restore is started, but only marked
databases are checked. If a configuration problem is detected, NMM stops the
restore operation.
7. Select the storage group linked to the RSG you want to recover. You can also
select individual mailbox databases in that storage group.
Only one storage group can be selected when “Restore to RSG” is enabled.
When Restore to RSG is enabled, Public Folders cannot be selected because
Exchange Server 2007 does not support restoring a public folder to the RSG.
Figure 38 on page 232 shows the entire SG2 storage group cannot be selected.
Only the Mailbox Database is selectable.
231
Figure 38
Public folders cannot be recovered to an RSG
Setting a different Exchange Server for recovery
NMM supports restoring to an RSG in a cluster environment. When an Exchange
server is set up in a cluster, and there are multiple active Exchange servers, you can
select the RSG you want to use for an Exchange recover session.
To select an active Exchange server RSG:
2. On the browse tab, right-click the Exchange server, and then click Change
Exchange Server.
Note: If the Exchange server is a stand-alone installation, this menu item will be
unavailable.
The Change Exchange Server dialog box appears.
3. In Available Exchange Servers, click the server you want to browse and then
click OK.
The browse tree is displayed with the RSG associated with the selected Exchange
server.
Setting Exchange Server 2007 permissions to enable NMM browsing of RSG mailboxes
An Exchange administrator must configure Exchange Server 2007 to enable browsing
of an RSG mailbox. If the Send-As and Receive-As permissions are not configured
correctly, if browsing the RSG is attempted, an error message dialog box will be
displayed stating that NMM is unable to browse mailboxes because the required
permission is not configured correctly.
To configure permissions for browsing an RSG mailbox:
1. On the Microsoft Exchange Server, select the Start > All Programs > Exchange
Management Shell.
2. Enable Send-As and Receive-As permission.
For example,
get-ExchangeServer <Exchange Server name> | Add-AdPermission -user
<username> -extendedrights Receive-As, Send-As
where:
• <Exchange Server name> is the server where the RSG mailbox is located.
232
• <username> is the name of the user being granted permissions.
Additional requirements for browsing RSG mailboxes
There are other requirements for browsing RSG mailboxes:
◆
An RSG must exist.
◆
The RSG must contain one or more databases.
◆
Databases must be online. NMM does not display mailboxes in offline databases.
◆
The MAPI/CDO kit must be installed, as described in “Microsoft Exchange
Server MAPI client and Collaboration Data Objects support” on page 194
◆
IPV6 must be disabled on Windows Server 2008, as described in “IPv6 issues
affect browsing RSG mailboxes” on page 220
◆
Exchange System Attendant and Information Store services must be running.
Selecting and recovering items in the RSG database
Once the databases have been recovered to an RSG, you can browse the user
mailboxes, folders, and messages, in the navigation tree, and then select which items
to recover to the user mailboxes. You can also search for items in the navigation tree
through the Search tab.
Selecting an item for recovery
Items can be selected for recovery at the storage group, mailbox, mail folder, and mail
message level.
To select items for recovery, click the checkbox beside a node in the navigation tree. A
check mark indicates that the node is selected.
To clear an item contained in a selected node, expand the node and click the checkbox
beside the item to clear it. The check mark disappears.
When you select a node, by default, all items contained in the node such as
mailboxes, mail folders, and messages, will also be selected for recovery. If all items in
the node are selected, the check mark will be green, and the background of the box
white.
If you want to select an item without selecting all items at the same level, expand
down to the level you want, without selecting. When you have located the item,
select the checkbox for the item. The containers above that item are automatically
partially selected. For example, in Figure 39 on page 234, the mail message “Is
everybody online?” has been selected for recovery, without selecting all items in the
containing mail folder, user mail box, or storage group. A gray check mark in a
checkbox with a gray background indicates that the item is partially selected. When
“Is everybody online” was selected, the nodes above were partially selected all the
way up to the server.
Figure 39 on page 234 depicts one selected mail message and several partially
selected nodes in the navigation tree.
233
Figure 39
Selected and partially selected items
Searching for an item
You can also search for a recovery item. Once you have located an item, you can select
it for recovery
To search for a recovery item:
1. Click Search above the navigation tree.
2. In the Path field, type a directory path.
3. Optionally, in the Name field, type the name of the search item. You can refine
your search as follows:
• Literal match (case-insensitive)
Type abc to return abc, ABC, or AbC but not abcd or ABCD.
• Literal match (case-sensitive)
Type "abc" to return abc but not ABc or abcd.
• Name contains (case-insensitive)
Type %abc% to return abc, abcd, ABCD, or xyzABCde.
• Name starts with (case-insensitive)
Type abc% to return abcd or ABCde but not xyzABCde.
• Name ends with (case-insensitive)
Type %abc to return xyzAbc but not ABCde.
• Single-character match search using the ? wildcard
Type ? to return single character entries and drive volumes, such as C or D.
Type WMI?Writer to return WMI Writer.
• Multiple-character match search by using the * wildcard
Type *.txt to return all entries with a .txt extension.
Type * to return all items within the selected container.
Type *writer* to return all writers.
234
• Search by using both the * and ? wildcard
Type *??I*writer* to return WMI Writer.
4. Click Search.
The Details pane displays the results of the search.
Recovering selected items
Once you have completed selecting items through browsing or searching, you can
recover the items to the users’ mailboxes.
To recover the selected items, from the Exchange Recover Session toolbar, click Start
Recovery.
Note: Recovered items do not overwrite existing items. They are placed in separate folders in
the user’s mailbox.
When NMM performs an RSG item level recovery, the items are placed in a new
“Recovered Items” folder created in the user’s mailbox on the production server.
Figure 40 on page 235 shows an example of two Recovered Items folders in the
mailbox of user Steve M.
Figure 40
Recovered items folders in user mailbox
The name of the folder begins with a time and date stamp. For example, “14:35:171
3/13/2009 Recovered Items.” A new Recovered Items folder is created for each
recovery. This prevents the selected items from overwriting the existing items in the
user’s mailbox, and allows the user to view and work with existing and recovered
items without overwriting either. When the user is done sorting through the
recovered items, the user can delete the Recovered Items folder.
Recovery to an alternate storage group
Recovery to an alternate storage group allows you to recover the storage group data
to a different Exchange server than the original source
Creating an alternate storage group and adding target databases
Before NMM can restore to an alternate storage group a new database must be
created in the alternate storage group, with the same name as the original database.
There are two methods available within Microsoft Exchange to configure an alternate
storage location database:
◆
Exchange Management Shell at command line. Recommended.
◆
Exchange Management Console Database Recovery Management tool.
235
To configure an alternate storage group with Exchange Management Shell:
1. On the Microsoft Exchange Server, select the Start menu and then select Exchange
Management Shell.
2. Use the new-storage group command to create the alternate storage group:
new-storagegroup –server exch_server –name alt_name
–LogFolderPath alt_path –SystemFolderPath alt_path
where:
• exch_server is the Exchange 2007 server name.
• alt_name is the desired alternate storage group name.
• alt_path is the path to the alternate storage group directory.
For example:
new-storagegroup –server bv-rmx64-1 –name Alternate_SG
–LogFolderPath T:\exchsrvr\SG2\Alternate_SG
–SystemFolderPath T:\exchsrvr\SG2\Alternate_SG
3. Use the new-mailboxdatabase command to add target databases to the alternate
storage group:
new-mailboxdatabase -Name db_name -MailboxDatabaseToRecover
db_name -StorageGroup exch_server\alt_name -EdbFilePath
edb_path
where:
• db_name is the name of the database that will be recovered.
• -EdbFilePath is the path of the database in the alternate storage group, not the
original storage group.
• edb_path is the path to the database file in the alternate storage group directory.
Note: The alternate storage group database name must match exactly the original database
name.
For example, the name for the alternate storage group specified in
new-mailboxdatabase -Name: is DB1, the same as the name of the original
database name specified in MailboxDatabaseToRecover, DB1:
new-mailboxdatabase -Name DB1 -MailboxDatabaseToRecover DB1
-StorageGroup bv-rmx64-1\Alternate_SG -EdbFilePath
U:\exchsrvr\SG2\Alterenate_SG\DB1.edb
4. Repeat the new-mailboxdatabase for each database that will recovered. Be sure
to specify the correct database name and edb path.
5. NMM allows the administrator to leave the database offline or have it
automatically put online after the restore. If you choose to leave the database
offline after the restore it will need to be manually mounted. This can be done
with the mount-database command:
mount-database –Identity exch_server\sg_name\db_name
For example:
mount-database –Identity bv-rmx64-1\Alternate_SG\DB1
Databases can be dismounted with the dismount-database command:
dismount-database –Identity exch_server\sg_name\db_name
236
For example:
dismount-database –Identity bv-rmx64-1\Alternate_SG\DB1
6. To repeat an alternate storage group restore by using the same database after it
has been mounted, set the database can be overwritten property:
set-mailboxdatabase –Identity
“exch_server\alt_sg_name\db_name” –AllowFileRestore:$true
For example:
set-mailboxdatabase –Identity “bv-rmx64-1\Alternate_SG\DB1”
To recover to an alternate storage group:
3. From the left pane, select Exchange Recover, and then select Database Recover.
5. Right-click the storage group that you want to recover, and then click Restore To.
6. In the Select Storage Group Restore Destination dialog box, select Alternate
Storage Group.
7. In the Exchange Server list, select the Exchange server that contains the alternate
storage group.
8. In Select Alternate Storage Group select the storage group, and then click OK.
237
Microsoft Exchange Server 2007 directed recovery
In the event of a problem recovering to the original Exchange server host, such as
disk drive failure, you can perform a directed recovery to recover Exchange data to a
different Exchange server.
The database portability feature in Microsoft Exchange Server 2007 allows a mailbox
to be mounted on any server in the same organization. “Database Portability” on
Microsoft TechNet at http://technet.microsoft.com/en-us/library/bb123954.aspx
describes this feature in more detail.
In the normal NMM backup and recovery process, the backup flows in the following
direction:
1. The NMM client is installed on the computer that will be backed up. For example,
ExchangeServer1.
2. That NMM client is configured to work with a specific NetWorker Server on
another computer. For example, NetWorkerServer1.
3. Backups are configured on and performed by the NetWorker Server, NetWorker
Server1.
4. Recovery is performed through the NMM client UI on the NMM client computer,
ExchangeServer1, and recovered to the same NMM client computer.
In a directed recovery, the NMM client is also installed on another server, for example
Server2. The recovery of ExchangeServer1 is “directed” from the NMM client on
Server2:
1. The NMM client is installed on the computer that will be backed up. For example,
ExchangeServer1.
2. The NMM client is also installed on another computer that will direct the
recovery. For example, Server2.
3. The NMM client on both machines are configured to work with the same
NetWorker Server on another computer. For example, NetWorkerServer1.
4. Backups are configured on and performed by the NetWorker Server, NetWorker
Server1.
5. Recovery of ExchangeServer1 is performed through the NMM client UI on another
NMM client computer, Server2, and recovered to another location.
To perform a directed recovery of an Exchange Server with NMM:
1. In NetWorker, create a backup of the NMM client, the Exchange Server.
2. Run NMM on the machine that will receive the data.
3. Connect to the NetWorker Server that hosts the NMM Exchange Server client.
4. Add the NMM Exchange Server client that created the backup as a locally
viewable client.
Figure 41 on page 239 shows the default of only one NMM client available, the
local client.
238
Figure 41
Local client on the taskbar next to Client
Figure 42 on page 239 shows the Configuration Options dialog box.
Refresh button
Figure 42
Configuration Options dialog box
Figure 43 on page 239 shows the Select Viewable Clients dialog box. The clients
available on the NetWorker server that you are attached to are listed under
Available clients on.
Figure 43
Select Viewable Clients dialog box
Microsoft Exchange Server 2007 directed recovery
239
7. Click the client to add in the Available clients on list box, and then click Add.
Add or remove clients as needed.
8. Click OK.
Figure 44 on page 240 shows the NMM window with the Client box.
Figure 44
NMM client dropdown list
9. Click the Client list, and select the client that created the Exchange Server backup
as the current local client.
11. On the Exchange Recover Sessions bar, select Recover Options.
The Exchange Recover Session Options dialog box appears.
12. Recover remote client's backup to local Exchange server.
240
Microsoft Exchange Server 2007 disaster recovery
NMM can perform the following types of disaster recovery for Microsoft Exchange
Server 2007:
◆
“Microsoft Exchange Server 2007 (standalone) disaster recovery” on page 241
◆
“Microsoft Exchange Server 2007 disaster recovery for Exchange CCR to a 2 node
cluster in a production environment” on page 242
◆
“Microsoft Exchange Server 2007 disaster recovery for Exchange CCR in a
production environment” on page 244
Microsoft Exchange Server 2007 (standalone) disaster recovery
The process for performing this disaster recovery follows these steps:
1. The Exchange data must have already been backed up with the save set
Applications:\Microsoft Exchange 2007.
2. Rebuild the Exchange Server hardware, following the instructions provided by
Microsoft.
3. After the Exchange Server is rebuilt, make sure with the name, account
information, and configuration, and other information is the same as the old
server that is being replaced.
4. Reinstall NMM.
5. In NMM, perform an Exchange Server point-in-time recovery.
The following procedure describes the process in more detail.
To perform a disaster recovery of a standalone installation of Microsoft Exchange
Server 2007:
1. Back up all Exchange data with the save set Applications:\Microsoft Exchange
2007.
2. Build new machine hardware for an Exchange Server, following Microsoft’s
instructions on TechNet:
“Moving Exchange Servers to New Hardware and Keeping the Same Server
Name”
http://technet.microsoft.com/en-us/library/bb332343.aspx
3. Capture all of your manually set Internet Information Services (IIS) virtual
directory configurations. Run the following Exchange Management Shell
command:
Get-OwaVirtualDirectory "owa (default web site)" |
export-clixml Owa.xml - depth
4. Copy the Owa.xml file to a location that can be accessed by the new server after it
is available and the old server is shut down.
5. Shut down the existing Exchange 2007 Server. (The backup should have already
been taken, at step 1 of this procedure.)
6. Reset the computer account for the existing Exchange 2007 server.
241
7. Bring the new computer online, and then confirm that the new computer is
running the same operating system that was installed on the existing Exchange
2007 server.
8. Rename the new server to the same name as the original server that you are
replacing, and then join this computer to the domain.
9. For drives that contained Exchange 2007 data, configure drive letters on the new
server to map to or match the configuration of the old server.
10. Verify that the drives have sufficient space to accommodate the restored data.
11. Run Exchange Server 2007 Setup with the following parameter:
Setup.com /M:RecoverServer
13. Restore your manually set IIS virtual directory configurations, by running the
following Exchange Management Shell script:
Restorevdir.ps1 Owa.xml
The Microsoft Exchange Server should come online with all previous storage
groups.
14. Install NMM.
The EMC Module for Microsoft Applications Release 2.2 Installation Guide provides
instructions for installing NMM, including Exchange Server-specific
configuration settings.
15. In NMM, perform an Exchange Server point-in-time recovery.
“Point-in-time recovery” on page 223 describes these recovery steps in detail.
Microsoft Exchange Server 2007 disaster recovery for Exchange CCR to a 2 node cluster in a
production environment
Prior to disaster recovery, use NMM to perform regular backups of volumes, System
Components, and application save sets.
C:\
System Components:\
• Application software
“Snapshot policies for SYSTEM COMPONENTS backup” on page 55 provides
more information about configuring snapshot policies for SYSTEM
COMPONENTS backup.
242
2. Back up application data for Exchange CCR on the NetWorker server on the
passive node using the following save set with snapshot policy "all."
Applications:\Microsoft Exchange 2007
The steps in this disaster recovery scenario are based on the following host details for
the production environment you are recovering, and the isolated environment you
are recovery to.
Production environment setup:
◆
Domain Controller, 1 physical machine
◆
CCR setup with an active node and a passive node, 2 physical machines
Isolated environment setup:
◆
Domain Controller setup with the same hardware and software configuration as
the production environment
◆
CCR setup with only a passive node, 1 physical machine
To perform a disaster recovery of Microsoft Exchange Server 2007 to a 2 node cluster
in a CCR environment:
1. Configure an AD domain setup in an isolated environment, using the same
operating system version, machine IP and machine name as the production
Active Directory Domain setup.
Note: There should not be any communication between the production environment and
the isolated environment.
2. Configure the AD domain on the isolated environment with the same domain
name as the production AD domain name.
Make sure that both environments use the same make and model of hardware.
3. Install all Windows updates, and applications such as Exchange, in the isolated
environment AD domain setup, to match the production Active Directory
Domain setup.
4. Disconnect the NetWorker server from the production environment, and connect
the NetWorker server to the isolated environment.
Note: In this scenario, the backup was taken on the file system, not on tape. Since the
backup is not on tape, to move the backup to the isolated environment you must connect
isolated environment to the NetWorker server which is storing the backup.
5. Install NMM on the isolated environment AD domain machine.
6. Restart the machine in "Directory Service Restore Mode" and start recovery of C:\
and System Components:\ using NMM recovery.
Note: Make sure that "legato" and "RMagentPS" folders are not selected before starting C:\
recovery.
7. When recovery completes, you will be prompted to restart the machine. Restart
the machine.
This recovers all production AD domain objects into the isolated AD domain.
This also recovers all application information.
243
8. On the isolated environment AD domain setup, reset the computer account for
the existing Exchange Virtual Server, or Clustered Mailbox Server (CMS), or
virtual Exchange Server.
9. Rebuild another standalone server in the isolated environment with the same
operation systemversion as the production CCR setup.
10. Provide this new machine with a new IP address and new Network name, and
bring it online in the isolated environment.
11. Install all roles required for Exchange setup to run.
12. Create the same drive letters and paths for Storage Groups (SG) and databases as
the production CCR SGs and databases.
13. Install the cluster feature and configure a new cluster with Quorum.
Note: The Majority Node Set (MNS)/Quorum shared folder must be different from the old.
14. Run Exchange setup in “Custom” mode, and only install the Passive Cluster Role.
15. Run the following command, using the CMS name and IP address from the
Production CCR server:
Setup.com /recoverCMS /CMSName:<name> /CMSIPaddress:<ip>
This puts the production virtual server online on this new node.
Note: All Storage Groups which were created in the CCR are recreated in this installation.
Verify that all drive letters and paths for Storage Groups and databases are recreated the
same as those in the production CCR server.
16. All databases will be in dismounted state, which is required for recovery from
NMM.
17. Install NMM on the new CCR machine.
18. Browse through Production CCR EVS index and perform PIT recovery of
Exchange.
19. Install Microsoft Exchange on the new CCR machine. Run Exchange setup in
Custom mode and install only the Passive Cluster Role.
20. Reseed the passive cluster node to get it in sync with the active node.
Microsoft Exchange Server 2007 disaster recovery for Exchange CCR in a production
environment
C:\
System Components:\
244
• Application software.
“Snapshot policies for SYSTEM COMPONENTS backup” on page 55 provides
more information about configuring snapshot policies for SYSTEM
COMPONENTS backup.
2. Back up application data for Exchange CCR on the NetWorker server on the
passive node using the following save set with snapshot policy "all."
Applications:\Microsoft Exchange 2007
The steps in this disaster recovery scenario are based on the following host details for
the production environment you are recovering, and the production environment
you are recovery to.
Production environment setup:
◆
Domain Controller, 1 physical machine
◆
CCR setup with an active node and a passive node, 2 physical machines
Recovery site setup (in the same network as the production setups):
◆
Domain Controller setup with the same hardware and software configuration as
the production environment
◆
CCR setup with only a passive node, 1 physical machine
In this disaster recovery procedure, the machines are referred to as follows:
Production:
◆
MC1: Domain controller
◆
CL1: cluster node 1
◆
CL2: cluster node 2
Recovery site (in the same network as the production setups):
◆
CL11: cluster node 1
◆
CL21: cluster node 2
To perform a Microsoft Exchange Server 2007 CCR disaster recovery in production to
a 2 node cluster:
1. Shut down the Exchange CCR setup.Shutdown both CCR nodes, assuming the
Exchange Cluster is lost or crashed.
2. Reset the computer account for the existing Exchange Virtual Server, or Clustered
Mailbox Server (CMS), or virtual Exchange Server.
3. Reset the computer account for CL1 and CL2 if you want the base cluster names
to be the same as those in original production setup.
4. Rebuild 2 machines CL21 and CL11 with the same operating system version as
the old server.
5. Assign the new machine with a new IP address and new Network name, and
bring it online in the same domain.
245
If you want the same IP address and Network name as the production machine,
assign the same IP and name.
6. Install all roles required for Exchange setup to run on both CL11 and CL21.
7. Install cluster feature and create a new cluster with Quorum
Note: The Majority Node Set (MNS)/Quorum shared folder must be different from the old.
8. Make sure the host has all of the drive letters created as those that were on the
original Exchange server.
9. Run Exchange setup in Custom mode and install only Passive Cluster Role on
CL11.
10. Run the following command, using the CMS name and IP address from the old
CCR server:
Setup.com /recoverCMS /CMSName:<name> /CMSIPaddress:<ip>
11. When this command is run:
• It puts the old virtual server online on this new node.
• All Storage Groups which were created in CCR are recreated in this
installation.
• All databases will be in a dismounted state, which is required to recover from
NMM.
12. Install NMM.
13. Browse through OLD CCR EVS index and perform PIT recovery of Exchange.
14. On CL211, install Exchange passive node. Run Exchange setup in custom mode
and install only the Passive Cluster Role.
246
9
Microsoft Data
Protection Manager
Backup
◆
◆
◆
◆
Data Protection Manager backup and recovery .....................................................
Performing DPM Server backups .............................................................................
Configuring a DPM Client resource .........................................................................
Performing DPM Server recovery ............................................................................
248
255
256
258
247
Data Protection Manager backup and recovery
NetWorker Module for Microsoft Applications (NMM) provides full backup and
recovery of Microsoft System Center Data Protection Manager (DPM) 2007. DPM is
Microsoft’s solution for backing up and recovering Microsoft application and system
data and files by using the VSS framework. DPM stores and manages this backup
information in its own database. DPM can provide backup and recovery for:
◆
Windows file services
◆
◆
◆
◆
Virtual Server 2005
NMM supports granular and directed granular recovery to DPM-protected clients.
This allows you to browse DPM replicas and select individual folders or files for
recovery. With directed recovery, data that originated on one NetWorker client host is
recovered to another NetWorker client host. Chapter 10, “Microsoft Data Protection
Manager Granular and Disaster Recovery,” provides more information about DPM
granular backup and recovery, and directed granular recovery.
Figure 45 on page 249 shows how NMM provides support for backup and recovery
of DPM data.
248
protects items within
DPM Server 2007:
DPM Configuration Database
DPM Replicas
NetWorker Server
7.3.3 or later:
Provides services for
NetWorker clients
such as NMM
DPM protects items within
Microsoft Servers:
Database files
Storage groups
Volumes
Transaction logs
LAN
Microsoft Servers:
Backup and Recovery
provided by DPM
Server 2007
DPM Server 2007:
Backup and Recovery
provided by NetWorker
Module for Microsoft
Applications (NMM) client
GEN-000754
Figure 45
NMM protection of DPM
backup procedures in Chapter 2, “Configuring a Scheduled Backup.”
DPM Server versions supported by NMM client
NMM client supports Microsoft DPM 2007 running on x86 and x64 machines.
249
DPM backup and recovery types
NMM supports several types of backup and recovery. For DPM, the NMM client
supports:
◆
DPM configuration database backup
Backup of the DPM configuration database
◆
DPM replica backup
Backup all or selected DPM Replicas for DPM-protected servers
◆
Backup level supported
Full only
◆
Recovery
Recovery of DPM Server can include recovery of the DPM configuration
database, or recovery of missing replicas in their entirety to their original
locations. NMM also supports granular file recoveries of replica data to the
DPM-protected client or to the DMP server.
DPM data backed up by NMM
NetWorker Module for Microsoft Applications supports only full backups for DPM
2007. It backs up the following DPM data:
◆
DPM configuration database: This is a SQL database. The DPM writer will
describe the location of the database, and NMM will use VSS SQL writers to
protect it.
Note: This version of NMM requires a local instance of SQL. It does not support remote
instances of SQL. The instance used by DPM Server must be on the same machine as DPM
Server.
◆
DPM replicas: These are data containers that DPM uses to store its save sets.
NMM will protect these as file system shares.
Note: A DPM replica is not a VSS replica, and is not directly manipulated outside of the
DPM environment.
Transportable backups of DPM replicas not supported
NMM 2.1 does not support transportable backup of DPM replicas. DPM replica is in a
dynamic volume. NMM does not support local nontransportable hardware-based
snapshots of dynamic volumes on Windows Server 2008. “Dynamic volume support”
on page 32 provides more information about dynamic volume and transportable
support in NMM 2.1
250
Components used by NMM for DPM backup and recovery
Table 38 on page 251 describes commands NMM uses to back up and recover DPM
data.
Table 38
NMM Data Protection Manager commands
Item
Description
Application writer
There is one writer for DPM 2007, “Microsoft DPM."
nsrsnap_vss_save
NMM command initiates the VSS-based backup.
nsrsnap_vss_recover
NMM command initiates the VSS-based recovery.
nsr_ps_single_save_per_sav
eset
NMM command specifies that PowerSnap create a separate snapshot for
each save set in the client entry. Whether the database should be online or
offline after restore. This attribute is required because of DPM writer
restrictions that require that the configuration database and each of the DPM
replicas must be separated into separate VSS shadow copy sets.
nsr_ps_first_saveset_save
NMM command specifies to save the DPM configuration database before
saving any of the DPM replicas. This is not required by the DPM writer., but is
recommended as a best practice by Microsoft.
dmpsync -restoredb -dbloc
[path] \DBMDB2007.mdf
DPM command instructing DPM to use the database at the specified location.
dpmsync -sync
DPM command instructing DPM to synchronize itself with the new
environment.
dpmsync -reallocatereplica
DPM command run on DPM Server before performing DPM replica recover.
This command instructs DPM Server to verify its replicas and to create the
appropriate number and size of partitions to receive the recovered replicas.
Note: For replicas using Custom Volumes, you cannot use this command. You
must manually create the necessary volumes before you can proceed with
DPM replica recovery. The Microsoft DPM documentation provides more
information about the use of Custom Volumes.
nsr_dpm_retry_wait
DPM command that specifies that if the DPM replica is unavailable for backup
because it is in use, wait the specified time (in minutes) before retrying to start
the backup. The default value is 10.
nsr_dpm_retry_maximum
DPM command that specifies the total number of times (he first try plus
retries) to attempt to start the backup. The default value is 3.
251
Specifying DPM save sets for application data
Table 39 on page 252 lists the DPM save set syntax to specify for supported types of
DPM data. Specify DPM data save sets in the Save Set attribute of a Client resource.
Some special characters in a DPM storage group name must be replaced with
URL-encoded values. “URL encoding for DPM save sets” on page 253 provides more
information about URL encoding.save set syntax:
Table 39
DPM Server 2007 save set syntax
Type of backup data
Save set syntax
Configuration database
There will always be exactly one DPM
configuration database.
APPLICATIONS:\Microsoft DPM\DPM
database.
Protection group replica set, of a non-VSS
datasource
There are usually multiple protection group
replica sets.
For example, a DPM replica resulting from
DPM protection of a system state on a server
named Mars.DPM.Server.
APPLICATIONS:\Microsoft DPM\Non VSS
Datasource Writer component
Computer%2FSystemState on server Mars
DPM Server
Protection group replica set
There are usually multiple protection group
replica sets.
For example, a DPM replica resulting from
DPM protection of a SQL database
Luna\FAcmeBank on a server named Pluto
SQL Server.
APPLICATIONS:\Microsoft
DPM\SqlServerWriter component
Luna%2FAcmeBank on server Pluto SQL
Server.
Displaying valid DPM data save sets
To display a list of the DPM save sets that are available for backup:
2. List the valid application data save set names.
nsrsnap_vss_save -?
Note: If the application server is on a virtual host, run the command from the physical
3. Press Enter.
APPLICATIONS:\Microsoft DPM\DPM database.
APPLICATIONS:\Microsoft DPM\Volume C: on server
mars.saturn.company.com
252
Turning the MSDE writer on or off in the registry
Although DPM Server may be backed up with either the MSDE writer or the SQL
writer, Microsoft recommends that you use the MSDE writer with DPM if it is
available. Windows Server 2008 does not include the MSDE writer.
To detect which writer is enabled, run nsrsnap_vss_save -? in a command window
on the DPM Server.
Among the writers displayed, you should see either a list of MSDE writer
components or a a list of SQL writer components.
On a Windows Server 2003 machine, you can enable the MSDE writer through the
following procedure:
To turn the MSDE writer on or off:
1. Open Registry Editor.
2. Locate the registry key:
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\VSS\Se
ttings\MSDEVersionChecking
3. To enable the MSDE writer, set the value to 0.
To disable the MSDE writer set the value to 1.
4. Exit Registry Editor.
5. Restart the VSS service.
URL encoding for DPM save sets
characters when naming DPM components:
The forward slash (/) in the component name Computer/SystemState is considered a
special character:
APPLICATIONS:\Microsoft DPM\Non VSS Datasource Writer component
Computer/SystemState on server Mars DPM Server
Replace it with its URL-encoded value, %2F:
APPLICATIONS:\Microsoft DPM\Non VSS Datasource Writer component
Computer%2FSystemState on server Mars DPM Server
values.
253
DPM application information variable settings
Table 40
DPM application information variable settings
Attribute name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
NSR_PS_SINGLE_SAVE_PE
R_SAVESET=value
Specifies to create a separate
PowerSnap save set for each save
set in the client entry.
yes
NSR_PS_SINGLE_SAVE_PER_SAVES
ET=yes
NSR_PS_FIRST_SAVESET_S Specifies the save set that
AVE=value
PowerSnap must operate upon first.
This must be set to the DPM
configuration database save set, as
Microsoft recommends that it be
saved first.
Configuration database save set name
For example:
NSR_PS_FIRST_SAVESET_SAVE=MY
_DPM_CONFIG_DB
NSR_DPM_RETRY_WAIT=val Specifies that if the DPM replica is
ue
unavailable for backup because it is
in use, wait the specified time (in
minutes) before retrying to start the
backup. The default setting is 1.
A positive integer
For example, to specify a wait time of two
minutes:
nsr_dpm_retry_wait=2
NSR_DPM_RETRY_MAXIMU
M=value
254
Specifies the total number of
times—the first try plus retries—to
attempt to start the backup.
A positive integer
For example, to specify of three attempts:
nsr_dpm_retry_maximum=3
Performing DPM Server backups
NMM DPM backup provides:
◆
◆
Full backup of DPM databases, logs, and storage groups, as well as
administrator-selected storage groups.
Configuring DPM backups
tasks in detail:
◆
◆
Note: For Data Protection Manager backups the Backup Snapshot policy must be set to All.
Backup will fail if any other value is used.
◆
◆
The next task is specific to configuring a DPM Client resource:
“Configuring a DPM Client resource” on page 256
Performing DPM Server backups
255
Configuring a DPM Client resource
policies.
You can create multiple Client resources for the same NMM host. In this way, you can
apply different backup attributes to different types of information on the same host.
For example, if the NMM host is an Exchange server, you can create one Client
resource to back up the Exchange databases and create another Client resource to
back up Windows system component data. This allows you to back up Exchange
databases many times a day and back up Windows system component data only once
a day.
Note: This version of NMM requires a local instance of SQL. It does not support remote
instances of SQL. The instance used by DPM Server must be on the same machine as DPM
Server.
To create a DPM Client resource:
Configuration.
To back up all DPM Server data, specify the save set name of every database and
replica. For example:
APPLICATIONS:\Microsoft DPM\DPM database.
APPLICATIONS:\Microsoft DPM\Volume H: on server
babaco.MARS.com
APPLICATIONS:\Microsoft DPM\Volume I: on server
babaco.MARS.com
“Displaying valid DPM data save sets” on page 252 describes how to find the
save sets available on the DPM Server. “Specifying DPM save sets for application
data” on page 252 describes the DPM Server save set syntax.
will be added.
256
If Client resources for the same NMM host are added to different backup groups,
ensure that the Start Time attribute for each backup group is spaced far enough
apart so that the backups for the host’s Client resources do not overlap.
13. In the Backup command attribute, type the backup command:
values, each on a separate line:
NSR_PS_SINGLE_SAVE_PER_SAVESET=yes
NSR_PS_FIRST_SAVESET_SAVE=APPLICATIONS:\Microsoft DPM\DPM
database
NSR_SNAP_TYPE=vss
“DPM application information variable settings” on page 254 provides
information about these variables and their values.
15. If a proxy client is being set up for the NMM, type the hostname of the proxy
client in the Remote Access attribute.
If the NMM is part of a cluster, type the names of the physical nodes of the cluster
in the Remote Access attribute.
18. Click OK.
Configuring a DPM Client resource
257
Performing DPM Server recovery
When the files are backed up, administrators can the perform the following types of
recovery:
258
◆
“Performing a granular recovery of DPM objects” on page 260
◆
“DPM disaster recovery” on page 267
10
Microsoft Data
Protection Manager
Granular and Disaster
Recovery
◆
◆
◆
Performing a granular recovery of DPM objects .................................................... 260
Performing a granular directed recovery to a DPM-protected server ................ 264
DPM disaster recovery ............................................................................................... 267
259
Performing a granular recovery of DPM objects
When NetWorker Module for Microsoft Applications (NMM) backs up a Microsoft
DPM Server, all of the DPM-protected server replicas that are backed up are indexed
to allow granular recovery of individual files in the replica. This granular recovery of
DPM replica files can be completed back to the DPM Server which is being protected,
or to the DPM-protected server. You can browse DPM replicas and select individual
folders or files for recovery. In DPM granular recovery, the DPM Configuration
database is not selectable.
For directed recovery to a DPM-protected server, NMM is installed on the DPM
Server for replica backup and on the DPM-protected server for directed recovery of
individual folders or files. This allows you to run the recovery of files in a
DPM-protected server replica, from the DPM Server, and direct the recovery to the
DPM- protected server computer.
Figure 46 on page 260 shows NMM set up for directed recovery of a DPM-protected
client server.
NetWorker Server:
Provides backup
services for
NetWorker clients
such as NMM
LAN
DPM-protected
servers: Microsoft
Applications
SharePointServer3
SQLServer4
FileServer1
Directed Recovery:
Granular file recoveries
to DPM-protected
servers, directly from
NetWorker Server,
through NMM client UI on
the DPM-protected server.
ExchangeServer2
DPM Server 2007:
NMM client provides
backup of DPM
Server and
DPM-protected
server replicas.
GEN-000958
Figure 46
Directed granular recovery to a DPM-protected client
“Performing a granular directed recovery to a DPM-protected server” on page 264
provides more information about using directed recovery together with DPM
granular recovery.
To recover a folder or file from a DPM replica:
3. Right-click the NetWorker server in the navigation tree.
When running NMM Recovery on a DPM Server, the default recovery mode is
Disaster Recovery. When running on a non-DPM Server, the default recovery
method is Granular, and the Disaster Recovery mode is unavailable.
260
Microsoft Data Protection Manager Granular and Disaster
Figure 47 on page 261 displays the DPM Recovery modes available.
Figure 47
DPM recovery modes
4. In the left pane, select Recover.
Figure 48 on page 261 displays the Recover submenu.
Figure 48
Recover submenu for DPM Recovery
5. Select DPM Recover Session.
6. Select Granular.
7. In the navigation tree, select the DPM files or folders to be recovered.
By default, the objects displayed in the navigation tree are from the most recent
backup.
Figure 49 on page 262 is an example of a DPM Server tree expanded to the file
level.
261
Figure 49
DPM granular recovery available to file level
8. Click Start Recover to begin the recovery operation.
Figure 50 on page 262 displays the options available for DPM Granular Recovery.
Figure 50
DPM granular recovery options
9. In the Recovery path box, specify the folder where the recovered items are to be
placed.
10. Select Display these options when initiating a Recovery to show this dialog box
when Start Recovery is selected. If this checkbox is cleared, options must be
through the Recover Options button on the main NMM screen.
11. Click OK.
15. Select Granular.
backup.
262
placed.
20. Click OK.
DPM Recovery Options Summary
This page lists the DPM Recovery and NetWorker Recover Options. This allows you
to review the settings before starting the recovery.
To change the NetWorker and DPM Recover Options, click Recover Options.
You can also access the settings from the DPM Recover Session view. On the
Exchange Recover Session toolbar click Recover Options.
The NetWorker Recover Options settings are specified on the General, NetWorker,
Security, and DPM tabs.
When the options are okay, click Start Recover to validate the options. If all option
settings are valid, the dialog box closes and recovery starts.
“Performing a granular recovery of DPM objects” on page 260 and “Performing a
granular directed recovery to a DPM-protected server” on page 264 provide more
information.
263
Performing a granular directed recovery to a DPM-protected server
In the basic NMM backup and recovery process, the backup flows in the following
direction:
1. The NMM client is installed on the computer that will be backed up.
2. That NMM client is configured to work with a specific NetWorker Server on
another computer.
3. Backups are configured on and performed by the NetWorker Server.
4. Recovery is performed through the NMM client UI on the NMM client computer,
and recovered to the same NMM client computer.
A more typical use for DPM backup would be a directed granular recovery. In the
NetWorker backup of DPM, NMM is installed on the DPM Server, to protect the DPM
Server, and the replicas of the DPM-protected servers. In DPM directed recovery, the
DPM Server is the NMM client that NetWorker backs up, and the recovery job is run
from the DPM Server, but the recovery is directed to the DPM-protected server.
DPM directed granular recovery scenario
The following scenario describes how DPM directed granular recovery would work
with a hypothetical FileServer1 as shown in Figure 51 on page 264.
NetWorker Server:
Provides backup
services for
NetWorker clients
such as NMM
LAN
DPM-protected
servers: Microsoft
Applications
SharePointServer3
SQLServer4
FileServer1
Directed Recovery:
Granular file recoveries
to DPM-protected
servers, directly from
NetWorker Server,
through NMM client UI on
the DPM-protected server.
ExchangeServer2
DPM Server 2007:
NMM client provides
backup of DPM
Server and
DPM-protected
server replicas.
GEN-000958
Figure 51
264
Directed recovery of FileServer1
Directed recovery in DPM scenario
In this scenario, the FileServer 1data was backed up through the NMM client on the
DPM Server, as part of the backup of DPM replicas. In the directed recovery, an NMM
client installed on FileServer 1 performs the recovery, instead of the NMM client on
the DPM Server:
1. The DPM Server protects a Windows Server Client, FileServer1.
2. The NMM client on the DPM Server backs up the DPM Server, including the
Protected-Client Replicas for the DPM Server, such as FileServer1.
3. The administrator receives a request to recovery a file, LostFile, from FileServer1.
4. The administrator checks DPM to see if DPM has the recoverable data.
DPM no longer has the file that needs to be recovered. But because NMM backs
up the file server replica as part of the DPM backup, the file should exist in a
previous NMM backup of DPM.
5. The NMM client is installed on FileServer1.
6. The NMM client on FileServer1 is configured for directed recovery from the DPM
Server.
Though the file the administrator wants to recover is from FileServer1, and it is
being recovered to FileServer1, the backup was performed through a remote client
on the DPM Server. The administrator must make the NMM client on the DPM
Server available for recovery through the NMM client on FileServer1.
7. In the NMM client on FileServer1 the administrator adds the DPM Server as an
available client.
8. Once DPM Server is an available client, the administrator can select items for
recovery from the remote client, the DPM Server, and recover them to the local
client, FileServer1.
Performing a granular directed recovery
To perform a granular directed recovery of a DPM -protected client:
Note: The backup must be a conventional backup, or a rolled over snapshot type backup.
1. Install NMM on the DPM-protected client, if it is not already installed.
The NMM client on the DPM Server, and the NMM client on the DPM-protected
client must be on the same NetWorker Server.
2. Open the NMM client on the DPM-protected client.
5. Click the client you want to add in the Available clients on list box, and then click
Add. Add or remove clients as needed.
6. Click OK.
7. Click the Client list box, and select the DPM Server NMM client that created the
backup as the current local client.
Performing a granular directed recovery to a DPM-protected server
265
9. On the System Recover Sessions bar, select Recover Options.
The NetWorker System Recover Session Options dialog box appears.
10. On the NetWorker tab, specify the destination for the recovery in the Relocate
Recovered Data box, and then click OK.
13. Select Granular.
backup. To recover objects from a previous backup:
placed.
18. Click OK.
266
DPM disaster recovery
In DPM disaster recovery, NMM allows the DPM database and replicas to be
selectable in the file view pane of NMM. Selecting replicas will select all missing
replicas, since they are not selectable individually.
Figure 52 on page 267 displays the DPM database and replicas in the file view pane.
Figure 52
DPM database and replicas available as selectable items
Note: Disaster recovery mode is not available when doing a directed recovery to a NetWorker
client other than the DPM Server which was originally backed up. When a remote client is
selected, this mode will be displayed as unavailable on the contextual menu.
You can browse DPM replicas and select individual folders or files for recovery. In
DPM granular recovery, the DPM configuration database is not selectable. In DPM
granular recovery, NMM directed recovery is available. “Performing a directed
recovery” on page 105 provides more information about NMM directed recovery.
To recover a folder or file from a DPM replica:
3. Right-click the NetWorker server in the navigation pane.
When running NMM Recovery on a DPM Server, the default recovery mode is
Disaster Recovery. When running on a non-DPM Server, the default recovery
method is Granular, and the Disaster Recovery mode is unavailable.
Figure 53 on page 267 displays the DPM Recovery modes available.
Figure 53
DPM recovery modes
Figure 54 on page 268 displays the Recover submenu.
267
Figure 54
Recover submenu for DPM recovery
6. In the navigation tree, select the DPM replicas or databases to be recovered.
backup.
7. Select Disaster.
Figure 55 on page 268 displays the options available for DPM disaster recovery.
Figure 55
DPM disaster recovery options
9. In the Recovery path box, specify the folder where the recovered databases are to
be placed.
10. Select Run DPM synchronization utilities for NMM to automatically run the
DPM sync commands after recovery.
11. Select Allocate non-custom volumes for replicas to specify for recovery to run
the dpmsync.exe to allocate disk volumes for any missing replicas prior to replica
recovery.
Note: This option does not apply to replicas created using "custom volumes." Custom
DPM volumes must be created manually. See Microsoft DPM documentation for more
information regarding custom volumes.
12. Select Initiate consistency checks on recovered replicas to specify for recovery to
initiate consistency checks on replicas following their recovery.
268
when “Start Recovery” is selected. If this checkbox is cleared, options must be
14. Click OK.
269
270
11
Microsoft Hyper-V
Backup and Recovery
◆
◆
◆
◆
◆
Hyper-V backup and recovery ..................................................................................
Performing Hyper-V backups ...................................................................................
Configuring a Hyper-V Client resource...................................................................
Performing Hyper-V recovery ..................................................................................
Hyper-V parent partition disaster recovery ............................................................
272
284
286
288
292
271
Hyper-V backup and recovery
NetWorker Module for Microsoft Applications (NMM) provides full backup and
recovery of Windows Server 2008 with Hyper-V and Windows Server 2008, Server
Core Installation, with Hyper-V.
Hyper-V is a Windows Server 2008 role that provides hypervisor-based server
virtualization. NMM utilizes the Hyper-V VSS writer to back up and recover the
following:
◆
Hyper-V Initial Store configuration file
◆
Each virtual machine, or child partition
backup procedures in Chapter 2, “Configuring a Scheduled Backup,” with the
specific details needed to back up and recover Hyper-V.
Hyper-V version requirements
NMM supports the Microsoft release-to-manufacturing (RTM) release of Hyper-V,
which is available only for the x64 edition of Windows Server 2008 RTM. The initial
Windows Server 2008 RTM provided a beta or release candidate (RC) version of
Hyper-V. If you are using one of these pre-RTM versions of Hyper-V, or did not order
Hyper-V with your Windows Server 2008 RTM license, you will need to download
the Hyper-V RTM update from Microsoft.
Operating system versions supported by NMM client
The NMM client supports Hyper-V parent partitions running Windows Server 2008
installations on 64-bit, virtualization aware, CPUs.
For parent partitions, NMM also supports Windows Server 2008 with the Server Core
installation.
When NMM is running in the Hyper-V parent partition, NMM can back up virtual
machines that are running the following guest operating systems:
◆
Windows Server 2008 (x86 and x64)
◆
Windows Server 2003 R2 (x86 and x64) SP2
◆
Windows Server 2003 (x86 and x64) SP2
◆
Windows Vista (x86 and x64) SP1
◆
Windows XP (x86) SP3
◆
SUSE Linux Enterprise Server 10 with SP1 x86 Edition
The NMM client also is supported in Hyper-V child partitions (guests). NMM
supports the same Windows Server operating systems and applications that are
supported on physical servers.
272
Hyper-V parent partition backup and recovery types
NMM provides these backup features:
◆
Backup types:
• Backup of Hyper-V server, child partitions, and applications within each child
partition
• Backup level: Full
◆
Backup granularity:
• Hyper-V backup on parent partition. All or each of the following:
– Hyper-V virtual machine child partitions
– Hyper-V Initial Store configuration file
NMM client provides these recovery features
◆
Recovery types:
• Conventional
• Restore from snapshot
• Directed Recovery:
– To the original machine
– To a different machine or location
– To a different Hyper-V server
• Disaster Recovery
◆
Recovery level:
• Full
• Hyper-V server (parent partition)
• Hyper-V role
• Hyper-V child partitions
• Individual Hyper-V child partitions
Hyper-V parent partitions, child partitions, and applications within child partitions
Hyper-V requires a parent partition that is running Windows Server 2008 to host the
child partitions. Each child partition is usually a server operating system running
applications, such as:
◆
◆
◆
◆
Hyper-V runs as a role in Windows Server 2008. NMM uses the Microsoft Hyper-V
VSS writer on the host to back up and recover Hyper-V through APPLICATION save
sets. The Hyper-V writer backs up and recovers Hyper-V configuration and virtual
machine files.
273
Figure 56 on page 274 illustrates a physical server running Windows Server 2008. The
Hyper-V role has been enabled on the server, and four virtual machines have been
created, each running a separate operating system and different Microsoft
applications.
Hyper-V virtual machines with child partition
operating systems and applications
Physical server with
Windows Server 2008
host operating system
and Hyper-V role
Virtual Machine VS-1
Windows Server 2008 x64
Configure Client resources within
each child partitions to back up
applications, volumes, and
system components on the
child partition
Windows Server 2003
Windows Server 2003
Windows Server 2003
Hyper-V configuration file
Configure Client resources
on parent partition to back
up parent partition volumes
and system components
Figure 56
Configure Client resources on parent partition to
back up Hyper-V child partitions and Hyper-V
configuration file
GEN-000959
Windows Server 2008 host with Hyper-V virtual machine child partitions
For complete data protection, configure NetWorker Client resources for each of the
following:
◆
The Hyper-V server, or parent partition: system components, volumes, and
applications
◆
Hyper-V virtual machine child partitions and configuration information on the
Hyper-V server
◆
The applications within each child partition
Hyper-V storage and backup options supported by NMM
There are a wide variety of storage configurations available for Hyper-V child
partitions, such as passthrough disks, direct-attached storage (DAS), Storage Area
Networks (SAN), and File Servers.
Figure 57 on page 275 illustrates Hyper-V storage options. Table 41 on page 275 lists
the corresponding configuration types by type of storage and connection.
274
The Microsoft website provides more details and the most up-to-date information
about storage hardware supported by Hyper-V.
up-to-date information about what software and hardware providers are supported
for Hyper-V backup and recovery in NMM.
Hyper-V Server
Parent Partition
Child Partition
X:\V1\V1.VHD
-> VHD1
VHD1 -> C:
DAS
Disk1-> X:
passthrough
Disk 2
Disk 2 -> D:
Configuration 1
Configuration 2
SAN
FC or iSCSI
Y:\V2\V2.VHD
-> VHD2
LUN 1 -> Y:
LUN1
FC or iSCSI
passthrough
LUN 2
LUN2
VHD2 -> E:
LUN2 -> F:
iSCSI
LUN 3 -> G:
LUN3
S:
SHARE1
CIFS/SMB/
SMBv2
\\FS\SHARE1
Configuration 3
Configuration 4
Configuration 5
Configuration 6
VHD3 -> H:
\\FS\SHARE1\V3\V3.VHD
-> VHD3
File Server
GEN-000962
Figure 57
Table 41
Hyper-V storage options
Hyper-V child partition configurations
Configuration
Type
1
VHD1 on DAS
2
DAS passthrough
3
VHD2 on LUN
4
LUN passthrough
5
iSCSI target attached within child partition
6
VHD3 on File Server
275
NMM supports Hyper-V snapshots of the child partitions and the parent partition
with the Microsoft Software VSS Provider or VSS Hardware Providers, depending on
the hardware storage type and partition type. The following list describes how these
configuration types are supported in NMM:
Snapshot support:
◆
Child partition: NMM is running within the child partition to perform the
backup. The following storage configurations are supported for this environment:
• If using the Microsoft Software VSS Provider, all configurations listed in
Table 41 on page 275 are supported.
• If using a VSS Hardware Provider, configuration 5 (LUN exposed directly to
child partition) is supported. Currently this support includes the EMC VSS
Hardware Provider with EMC CLARiiON storage. Check the EMC Information
Protection Software Compatibility Guide for the latest support information.
◆
Parent partition: NMM is run in the parent to perform the backup. The following
storage configurations are supported for this environment:
• If using the Microsoft Software VSS Provider, configuration 1 (VHD1) and
configuration 3 (VHD2) are supported.
• If using a VSS Hardware Provider, configuration 3 (VHD2) is supported.
Currently this support includes the EMC VSS Hardware Provider with EMC
CLARiiON storage. Check the EMC Information Protection Software
Compatibility Guide for the latest support information.
Note: When performing child partition backups while executing on the parent
partition, the Microsoft Hyper-V Writer does not include the passthrough or
child-attached iSCSI drives for a virtual machine. As such, configurations 2, 4, and 5
are not supported by the Hyper-V Writer. Configuration 6 is not supported because the
VSS framework does not support network shares.
Cluster support: In addition to the supported configurations listed for snapshots,
parent and child cluster scenarios support the following storage configurations.
◆
Child partition clustering: NMM is running within the child partition to perform
the backup:
• For drives configured as failover resources, configuration 5 (LUN exposed
directly to child partition) is supported as this is the only configuration
Microsoft currently supports for Windows Server 2008 Failover Clustering.
• For drives configured for operating system or local drives for the cluster
nodes, the child partition support listed under Snapshot support applies.
◆
Parent partition clustering: NMM is run in the parent to perform the backup:
• All configurations are supported.
276
Backup roadmap for Hyper-V
Table 42 on page 277 describes the backup tasks for the Hyper-V parent and child
partitions.
Table 42
Backup tasks for Hyper-V (page 1 of 2)
Hyper-V parent partition
Tasks
On the parent partition
Windows volumes:
• Volume data such as a drive letter, for
example D:\.
• A file system path, such as D:\data.
Windows system components, such as:
• The registration database
• Windows Cluster, etc.
• Other Window system components,
as listed in the Table 1 on page 31.
Complete tasks 1 through 7:
page 52
277
Table 42
278
Backup tasks for Hyper-V (page 2 of 2)
Tasks
On the parent partition
The Hyper-V role can coexist with other
applications, such as:
• Microsoft Office SharePoint Server
• Microsoft System Center Data
Protection Manager (DPM)
lists the specific versions supported for
each of these applications.
Configure application backups with the specific instructions for the
application in:
page 52
“Task 7: Configure a proxy client” on page 67“Performing SQL
Server backups” on page 119
Hyper-V on parent partition
Hyper-V virtual machines and Initial Store
configuration file
Complete tasks 1 through 4:
page 52
• Review the information in:
“Best practices for Hyper-V parent partition backup and recovery”
on page 279
“NMM Hyper-V commands” on page 281
• Complete the tasks in:
“Configuring a Hyper-V Client resource” on page 286
Hyper-V child partition applications
Application data, such as:
• Microsoft Office SharePoint Server
• Microsoft System Center Data
Protection Manager (DPM)
• Microsoft Hyper-V
lists the specific versions supported for
each of these applications.
Configure application backups with the specific instructions for the
application in:
Best practices for Hyper-V parent partition backup and recovery
This type of backup uses the Hyper-V writer on the parent partition:
◆
To get the most benefit from the Hyper-V role, create separate child partitions for
each application, so that the only application-type backup and recovery
performed at the host level is for Hyper-V.
◆
After disaster recovery of the parent partition, you may need to recover
applications within each child partition if you are doing separate child partition
backups, and those backups are more recent than the complete parent partition
backups. “Hyper-V parent partition disaster recovery” on page 292 provides a
complete list of these tasks.
◆
Best use for this type of backup is bare metal recover of a guest and for recovery
of operating system roles.
◆
Best practice for Initial Store backup is to back up when Hyper-V configuration
changes are made. Initial Store does not need to be backed up each time a virtual
machine guest is backed up.
◆
The Hyper-V writer does not support backup of the configuration file Initial Store
to a proxy client.
◆
The primary purpose for restoring Initial Store in NMM is for disaster recovery of
the Hyper-V server.
◆
Back up Data Protection Manager, Exchange, Microsoft SQL Server, or Microsoft
SharePoint applications from within the child partition. The Hyper-V backup is
not recommended as the method to back up applications on a child partition. In
Hyper-V backup:
• The child partition copy backup method is used.
• There is no child partition log management. For example, Exchange logs are
not truncated.
◆
Roll-forward recovery is not available for virtual machine/child partition level
disaster scenarios. From a parent partition, a roll-forward recovery of a child
partition is not possible. Recoveries from a parent partition are point-in-time
(disaster recovery).
◆
Child partition pass-through disks are skipped during Hyper-V parent partition
backup.
◆
Basic disks only are supported only within child partitions.
Backup of dynamic disks within child partitions is not supported. The guest
snapshot is mounted during the Hyper-V backup process, and this changes the
disk signature on dynamic disks in a guest.
◆
Any VSS Hardware or Software Provider can be used. The Hardware Provider
must support Windows 2008 extensions for AutoRecoverable snapshots.
◆
Windows Server 2008 Failover Clustering is supported—allows you to configure
failover of virtual machines.
◆
EMC Storage connected to Fibre Channel or iSCSI storage can be used in the
parent to host virtual machines.
◆
Do not take a Hyper-V VSS parent partition snapshot of Hyper-V child partitions
that are part of a SharePoint farm. To back up SharePoint on the Hyper-V child
partition, install the NMM client on the child partition, and perform the Share
Point backup locally from within the child partition.
279
For more information about using SharePoint and Hyper-V together, check the
Microsoft website for recommendations and requirements.
Best practices for Hyper-V child partition applications backup and recovery
This type of backup and recovery is performed within the child partition, and uses
application and system components writers available on that child partition:
◆
The Microsoft Exchange team recommends using backups within the child
partition as the preferred method for Exchange backup and recovery.
◆
NMM supports roll-forward recovery for Microsoft Exchange, when Exchange is
backed up within the child partition.
◆
Within child partitions, standard application backup and recovery rules and
capabilities apply, including roll-forward recoveries.
◆
VSS Hardware Providers for iSCSI storage are supported for iSCSI disks that are
native within the child partition.
◆
VSS Hardware Providers for Fibre Channel storage are supported when the
Hardware Provider can support hardware snapshots without Custom
ControlDataBlock (CDB):
• CLARiiON arrays are supported because they do not require custom CDB.
• Symmetrix/DMX arrays are not supported because they require custom CDB.
• For VSS Hardware Provider, a transportable snapshot must be created, and the
proxy node must be a physical machine.
• The proxy cannot be the host machine, and it must match the operating system
version of the child partition. This requirement is independent of Hyper-V.
◆
Child partition pass-through disks are skipped in Hyper-V parent partition
backup; child partition pass-through disks are supported by backups within the
child partition.
◆
Windows Server 2008 Failover Clustering with iSCSI storage is supported.
◆
Child partition Windows Server 2008 Failover Clustering with Fibre Channel
storage is not supported because SCSI-3 is not supported in Hyper-V child
partitions.
◆
Windows Server 2003 Clustering is supported, but at the time of this NMM 2.2
release, Microsoft has not issued a support statement for it.
Restrictions on relocating and restoring to other locations
Hyper-V has several restrictions on relocating and restoring to other locations:
280
◆
Directed recover or relocated recover of Hyper-V objects on a Hyper-V Server
Cluster is not supported.
◆
Restoring Hyper-V VMs to non-Hyper-V Servers is not supported.
◆
Relocation recover of Initial Store is not supported.
◆
Relocation or redirection of NMM Hyper-V back-ups taken before a NMM2.2
upgrade is not supported.
◆
When a VM is redirect recovered to a second Hyper-V Server, the user must
update the Network Adapter settings of the VM with the Hyper-V Manager
before starting the VM.
Components used by NMM for Hyper-V backup and recovery
Table 43 on page 281 describes commands NMM uses to back up and recover
Hyper-V data.
.
Table 43
NMM Hyper-V commands
Item
Description
Application writer
There is one VSS writer for Hyper-V, “Microsoft Hyper-V."
nsrsnap_vss_save
NMM command that initiates the VSS-based backup.
nsrsnap_vss_recover
NMM command that initiates the VSS-based recovery.
Specifying Hyper-V save sets for application data
Figure 58 on page 281 describes what the NMM client backs up in Hyper-V, using the
Microsoft Hyper-V VSS writer and NMM save sets.
“APPLICATIONS:\Microsoft Hyper-V”
Save set backs up Hyper-V components, which includes
each virtual machine and the Hyper-V configuration file
Parent Partition:
Windows Server 2008
OS with Hyper-V Role
“APPLICATIONS:\Microsoft Hyper-V\VS-1”
Save set to back up virtual machine “VS-1”
VS-1
Virtual Machine
Child Partition
“APPLICATIONS:\Microsoft Hyper-V\VS-2”
Save set to back up virtual machine “VS-2”
VS-2
Virtual Machine
Child Partition
“APPLICATIONS:\Microsoft Hyper-V\Initial Store”
Save set to back up configuration file “Initial Store”
Hyper-V Configuration File
GEN-000960
Figure 58
NMM backup of Hyper-V components
281
Table 44 on page 282 lists the Hyper-V save set syntax to specify for supported types
of Hyper-V backup. Specify Hyper-V data save sets in the Save Set attribute of a
Client resource.
Table 44
Hyper-V save set syntax
Type of backup data
Save set syntax
Hyper-V Manager
The entire Hyper-V configuration: the
Hyper-V configuration file, and each virtual
machine
APPLICATIONS:\Microsoft Hyper-V
Hyper-V configuration file, also known as
“Initial Store.” There is one configuration file
in the Hyper-V Manager installation, listing
the Hyper-V settings for the host operating
system, and the guest operating systems.
Hyper-V\Initial Store
Hyper-V virtual machine. There are usually
multiple virtual machines on the host
operating system.
Note: The Hyper-V writer does not support offline backup of the
configuration file. The ”APPLICATIONS:\Microsoft
Hyper-V“ save set cannot be used in a proxy backup group.
Note: The Hyper-V writer does not support offline backup of the
configuration file. The ”APPLICATIONS:\Microsoft
Hyper-V\Initial Store save set cannot be used in a
proxy backup group.
Hyper-V\virtual_machine_name
Virtual machines can be included in a proxy backup group.
Displaying valid Hyper-V save sets
To display a list of the Hyper-V save sets that are available for backup:
1. Open a command prompt on the Hyper-V server.
2. List the valid application data save set names:
• If the Hyper-V server is not on a cluster virtual host, type:
nsrsnap_vss_save -?
• If the Hyper-V server is on a cluster virtual host, type:
nsrsnap_vss_save -? -c cluster_virtual_server_name
Note: If the application server is on a cluster virtual host, run the command from the
physical node that is currently hosting the application server.
3. Press Enter.
The following example shows the application data save sets that are available on
a Hyper-V system with two virtual machines, virtual_machine_name_1 and
virtual_machine_name_2:
Hyper-V
Hyper-V\Initial Store
Hyper-V\virtual_machine_name_1
Hyper-V\virtual_machine_name_2
The list of Microsoft Hyper-V save sets will include the top-level entry for
Microsoft Hyper-V, the configuration file, and all the virtual machines if any exist.
282
Hyper-V application information variable settings
Table 45
Hyper-V application information variable settings
Attribute name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
283
Performing Hyper-V backups
NMM Hyper-V backup provides full backup of Hyper-V virtual machines and
Hyper-V Initial Store configuration file.
Preparing a virtual machine for backup
Before NMM can create a backup of a Hyper-V virtual machine, Integration Services
must be installed on the virtual machine. These services provide integration between
the physical server and the virtual machines, such as time synchronization, guest
operating system shutdown messaging, and backup support.
Integration Services can only be installed after the guest operating system has been
installed on the virtual machine.
To install Integration Services, on the virtual machine guest system click the Action
menu, click Insert Integration Services Setup Disk, and then click the .exe to run
setup.
When Windows Server 2008 is installed as the operating system on a guest system,
Integration Services are installed by default.
Hyper-V configuration requirements for backing up a virtual machine that contains multiple
volumes
When there are multiple virtual hard disks in the guest, backup of the associated VM
from the parent partition may fail because of a Microsoft limitation. When there are
multiple volumes on the guest, VSS determines the shadowstorage area for the
snapshots based on which volume has more space. This can lead to a condition where
volume C’s snapshot and volume D’s snapshot both reside on volume D since
volume D has more space. During the snapshot revert stage in PostSnapshot, volume
C’s snapshot may be lost if volume D’s snapshot is reverted first.
To prepare a multiple volume guest for backup, use the vssadmin command to force
the shadowstorage of each volume to be on the same volume:
Note: These commands must be run inside each guest, not the parent physical Hyper-V Server.
vssadmin Add ShadowStorage /For=C: /On=C:
vssadmin Add ShadowStorage /For=D: /On=D:
Repeat as needed for each volume in the VM.
NMM support for proxy backup of Hyper-V
NMM does not support Storage Area Network (SAN) boot configurations, where a
hardware snapshot of the system drive is performed. The Microsoft Hyper-V writer
requires that the configuration component “Initial Store” always reside on the system
partition. This path cannot be changed.
Virtual machine-only save sets can be specified in a proxy backup group:
APPLICATIONS:\Microsoft Hyper-V\virtual_machine_name_1
284
The Microsoft Hyper-V save set and configuration file (Initial Store) save set cannot
be specified in a proxy backup group:
APPLICATIONS:\Microsoft Hyper-V\Initial Store
Configuring Hyper-V backups
tasks in detail:
◆
◆
Note: Hyper-V Virtual Machine snapshots are not related to NMM or NetWorker
snapshots. Hyper-V Virtual Machine snapshots are created, viewed, and applied to the
virtual machine through Hyper-V Manager. When NMM backs up a Hyper-V virtual
machine, the Hyper-V Virtual Machine snapshots are part of that backup.
◆
◆
The next task replaces the general instructions for Task 5 with information specific to
configuring a Hyper-V Client resource:
“Configuring a Hyper-V Client resource” on page 286
Performing Hyper-V backups
285
Configuring a Hyper-V Client resource
policies.
You can create multiple Client resources for the same NMM host. In this way, you can
apply different backup attributes to different types of information on the same host.
For example, if the NMM host is an Exchange server, you can create one Client
resource to back up the Exchange databases and create another Client resource to
back up Windows system component data. This allows you to back up Exchange
databases many times a day and back up Windows system component data only once
a day.
To create a Hyper-V Client resource:
Configuration.
To back up all of Hyper-V—the virtual machines and the configuration file (Initial
Store)—specify:
To back up the virtual machines, specify:
To back up the configuration file (Initial Store), specify:
Note: The Microsoft Hyper-V save set and configuration file (Initial Store) save set cannot
be specified in a proxy backup group:
286
“Displaying valid Hyper-V save sets” on page 282 describes how to find the save
sets available on the Hyper-V host. “Specifying Hyper-V save sets for application
data” on page 281 describes the Hyper-V save set syntax.
will be added.
If Client resources for the same NMM host are added to different backup groups,
ensure that the Start Time attribute for each backup group is spaced far enough
apart so that the backups for the host’s Client resources do not overlap.
14. In the Backup command attribute, type the backup command:
NSR_SNAP_TYPE=vss
NSR_DATA_MOVER=<proxy_client_name>
Type application variable settings and their values on separate lines.
16. If a proxy client is being set up for the NMM client, type the hostname of the
proxy client in the Remote Access attribute.
If the NMM is part of a cluster, type the names of the physical nodes of the cluster
in the Remote Access attribute.
19. Click OK.
Configuring a Hyper-V Client resource
287
Performing Hyper-V recovery
When the files are backed up, administrators can recover all of the Hyper-V
components, the Initial Store configuration file, or individual virtual machines,
depending on what was specified in the backup save set.
The following sections describe Hyper-V recovery:
◆
“Off-line and on-line recovery” on page 288
◆
“Hyper-V recovery to the original machine and location” on page 289
◆
“Hyper-V recovery to a different machine or location” on page 289
Off-line and on-line recovery
A virtual machine must be offline before recovery can start. If the virtual machine is
online at the time recovery is started, the Hyper-V writer will first turn off the virtual
machine.
Once the virtual machine is offline, the current virtual machine is destroyed, the
recovery version is restored, and the virtual machine is registered.
The Hyper-V writer automatically handles detecting whether the virtual machine is
on-line, and turning it off. There is no action required by the administrator.
Restrictions for backup and recover of Hyper-V virtual machines in a Windows 2008 Failover
Cluster
To back up a Hyper-V virtual machine that resides on a physical machine which is
part of a Windows 2008 Failover Cluster, you cannot backup or recover the virtual
machine as part of the cluster virtual server.
For example, consider the following Failover Cluster setup:
◆
A cluster, Cluster_Virtual_Name, contains two physical machines,
Physical_Machine_1 and Physical_Machine_2.
◆
Physical_Machine_1 contains two virtual machines, VM1 and VM2.
◆
You want to back up and recover VM1.
If you create a NetWorker client resource for Cluster_Virtual_Name, backup and
recover of VM1 through that client resource is not supported.
If you create a client resource for Physical_Machine_1 you can specify to back up the
Hyper-V Application (APPLICATIONS:\Microsoft Hyper-V) which will include all
virtual machines on the physical machine, or you can specify an individual virtual
machine VM1 (APPLICATIONS:\Microsoft Hyper-V\VM1).
You can recover VM1 from the NW client resource of the physical machine,
Physical_Machine_1. This type of recovery is described in “Hyper-V recovery to the
original machine and location” on page 289.
You can use redirected recover to recover VM1 from the NW client resource of the
physical machine, Physical_Machine_1 to the Physical_Machine_2. This type of
recovery is described in “Hyper-V recovery to a different machine or location” on
page 289.
288
Hyper-V recovery to the original machine and location
To recover Hyper-V components to the original machine and location:
1. Open the NMM software and select the NetWorker server on which the NMM
software was configured for backup.
If NMM is part of a cluster, select the physical client to which you are recovering
data. The physical client can be selected from the Client list attribute in the
application toolbar.
2. From the left pane, select Recover > Hyper-V Recover Session.
3. From the navigation tree, select the Microsoft Hyper-V writer, or individual
virtual machines under the Microsoft Hyper-V writer.
4. From the Hyper-V Recover Session toolbar, click Recover.
The Hyper-V Recover Session Summary dialog box is displayed.
5. If all the options look okay, go to step 11 .
or
If you want to change options, go to step 7 .
6. Click Recover Options.
7. On the General tab, specify the Diagnostic Output Level.
8. On the NetWorker tab specify the Restore Type, and select or clear Terminate
recover of item if errors are encountered.
Note: Restore of Hyper-V components requires a Restore Type of Conventional Restore.
9. On the Security tab, specify pass phrases if any are needed.
10. Click OK to close the Hyper-V Recover Session Options dialog box.
NMM validates all pages. If all pages are valid, recovery begins.
Hyper-V recovery to a different machine or location
To recover Hyper-V components to a different machine or location:
1. Open the NMM software and select the NetWorker server on which the NMM
software was configured for backup.
If NMM is part of a cluster, select the physical client to which you are recovering
data. The physical client can be selected from the Client list attribute in the
application toolbar.
2. From the left pane, select Recover > Hyper-V Recover Session.
3. From the navigation tree, select the Microsoft Hyper-V writer, or individual
virtual machines under the Microsoft Hyper-V writer.
4. From the Hyper-V Recover Session toolbar, click Advanced Recovery.
The Hyper-V Restore Wizard starts and the Destination Host page is displayed.
289
5. Specify the destination host server for the Virtual System recover:
• To recover to the same location as the original, select Recover (Overwrite)
Virtual System to original location, and then go to “To perform validation
and start recovery:” on page 291.
• To recover to a different path on the same Hyper-V server:
a. Select Recover Virtual System to a different path and then click Next.
The Destination Path page is displayed, and you can specify a destination
for each Virtual System.
b. Click Browse to specify the destination location for the configuration files.
After selecting the destination location for the configuration files, you can
change the destination location for the virtual machine’s virtual disks.
c. To change the destination location for a virtual disk, click the virtual
machine’s virtual disk in the list, and then click Change Destination.
Repeat as needed for each virtual disk destination that you want to change.
d. When you have completed changing destinations, go to “To perform
validation and start recovery:” on page 291.
Note: The destinations provided on this page are Microsoft's default configuration
file locations and may not match your Hyper-V configuration. Change the
destination as needed.
• To recover to a different Hyper-V server:
a. Click Recover Virtual System to a different Hyper-V server.
b. From the Select Remote Host list click the server you want to recover to.
c. Click Next.
The Destination Path page is displayed, where you can specify a
destination for each Virtual System.
d. Click Browse to specify the destination location for the configuration files.
After selecting the destination location for the configuration files, you can
change the destination location for the virtual machine’s virtual disks.
e. To change the destination location for a virtual disk, click the virutal
machine’s virtual disk in the list, and then click Change Destination.
Repeat as needed for each virtual disk destination that you want to change.
f. When you have completed changing destinations, go to “To perform
validation and start recovery:” on page 291.
Note: The destinations provided on this page are Microsoft's default configuration
file locations and may not match your Hyper-V configuration. Change the
destination as needed.
290
To perform validation and start recovery:
1. Click Finish.
NMM performs validation. If the validation is not successful, an error message is
displayed. If the validation is successful, a Summary page is displayed listing the
Hyper-V and NetWorker Recover options specified. If you need to change any of
the options, click the Recover Options or Back button.
NMM validates all pages. If all pages are valid, recovery begins.
291
Hyper-V parent partition disaster recovery
Successful disaster recovery starts with good backups at all levels on the Hyper-V
physical server. The frequency of these backups depends on your retention policies,
the frequency that the data on your servers and applications changes, and the type of
application. Regardless of the frequency of your backups, disaster recovery must
proceed in a specific order starting with the host machine and ending with the
application data on your virtual machine guests. Figure 59 on page 292 describes the
order and types of recovery.
Hyper-V virtual machines with guest
operating systems and applications
Physical server with
Windows Server 2008
host operating system
Windows Server 2008 x64
4. On each guest, recover the
applications using the recovery
process appropriate for each
application.
Windows Server 2003
Windows Server 2003
Windows Server 2003
Hyper-V configuration file
1. On the physical server
host, recover the host
volumes and system
components.
2. Re-enable the Hyper-V
role.
Figure 59
3. On the physical server host, recover the
Hyper-V virtual machine guests and
Hyper-V configuration file.
GEN-000961
Disaster recovery order for a Hyper-V server
Full disaster recovery requires that the following backups are performed ahead of
time on a regular schedule:
◆
Physical host:
• Application data
• Windows volumes
• Windows system components
292
◆
Hyper-V on host: Hyper-V virtual machines and Initial Store configuration file
◆
Hyper-V in-guest: Application data
!
IMPORTANT
Before starting disaster recovery, make sure the target machine is up to date with
the “Hyper-V Update for Windows Server 2008 x64 Edition (KB950050).” This
article and download is available from the Microsoft Download Center. If this
update is not installed prior to recovery, the system will not come back after
recovery of SYSTEM COMPONENTS.
To perform a Hyper-V host disaster recovery:
1. Review best practices:
• “Best practices for Hyper-V parent partition backup and recovery” on
page 279
• “Best practices for Hyper-V child partition applications backup and recovery”
on page 280
2. Recover/restore the physical host which was backed up by NMM.
“Disaster recovery” on page 109 provides more information about recovering the
NMM client.
3. Re-enable the Hyper-V role on the host.
Consult Microsoft documentation for information about enabling the Hyper-V
role on Windows Server 2008.
4. On the host, perform a recovery of Hyper-V as described in “Performing Hyper-V
5. From within each guest, perform a recovery of each application, using the
appropriate procedure for the application.
“Performing a recovery” on page 86 provides more information about recovering
applications, including links to specific recovery steps for each application.
Hyper-V parent partition disaster recovery
293
294
12
Microsoft Windows
Cluster Backup and
Recovery
◆
◆
◆
◆
Windows Server Cluster backup and recovery ......................................................
Performing Windows Server Cluster backups........................................................
Windows Server Cluster recovery ............................................................................
Windows Server 2003 clustered client disaster recovery ......................................
296
298
299
299
295
Windows Server Cluster backup and recovery
(NMM) backup procedures in Chapter 2, “Configuring a Scheduled Backup,” and
recovery procedures in Chapter 4, “System Backup and Recovery,” with the specific
details needed to back up and recover clusters in:
◆
Microsoft Windows Server 2008 — Windows Server Failover Clustering (WSFC)
◆
Microsoft Windows Server 2003 — Microsoft Cluster Service (MSCS)
Cluster support in NMM client
NMM client supports Windows Server Clusters running on the following versions of
Microsoft Windows Server:
Microsoft Cluster Service (MSCS)
Microsoft Windows Server 2003 SP2 (x86, x64) and Microsoft Windows Server 2003
R2 with SP2 (x86, x64):
◆
Standard Edition
◆
Enterprise Edition
◆
Datacenter Edition
Windows Server Failover Clustering (WSFC)
Microsoft Windows Server 2008 (x86, x64):
◆
Standard Edition
◆
Enterprise Edition
◆
Datacenter Edition
Compatibility with previous Windows Cluster backups
Windows Server 2008 does not support recoveries of Windows Server 2003 backups
of the Cluster Writer. Many Windows legacy APIs in Windows Server 2003 are not
supported in Windows Server 2008. NMM client does not prevent attempting to
recover a Windows Server 2003 Cluster to a Windows Server 2008 Cluster, but
recovery will fail.
Components used by NMM for Windows Cluster backup and recovery
NetWorker Module for Microsoft Applications backs up Windows Cluster data as
part of the SYSTEM COMPONENTS save set. There is a Cluster Writer used within
SYSTEM COMPONENTS, but it cannot be specified in a save set.
The Cluster writer is recovered when one of the following is selected for recover:
296
◆
SYSTEM COMPONENTS — As part of recovery of this entire save set
◆
SYSTEM COMPONENTS:\Cluster Writer — As an authoritative restore in
Windows Server 2008
◆
SYSTEM COMPONENTS:\Cluster Writer and SYSTEM
COMPONENTS:\BootableSystemState — As part of a nonauthoritative restore in
Windows Server 2008
Windows Cluster application information variable settings
Table 46 on page 297 lists the variables that can be specified in NetWorker
Management Console, in the Application information attribute of the Client resource.
Table 46
Windows Cluster application information variable settings
Attribute Name
Description
Values
NSR_SNAP_TYPE=value
provider name.
vss
NSR_DATA_MOVER=valuea
pplication information attribute
client that moves snapshots from
• The local host
client.
If setting up an Windows Cluster client
specified, Windows Cluster consistency
checks are performed on the proxy client.
Related messages are written to the
Replication Manager log files
(erm_clientxx.log) on the proxy client.
Windows Cluster client resource,
client resource host. Related messages
are logged on the client resource host in
both the Replication Manager log files
and the nmm.raw file.
Windows Server Cluster backup and recovery
297
Performing Windows Server Cluster backups
To configure a virtual NMM client for backup, complete all of the tasks required to
back up a regular NMM client. There are additional considerations for setting up a
virtual NMM client, which are outlined in Table 47 on page 298.
Table 47
Additional steps and considerations for backing up a clustered Client resource
Backup task
Consideration
“Task 4: Configure a backup
group”
Configure a Client resource for each virtual server that is being backed up and
each physical node in the cluster on which the virtual server may execute.
Type the names of the physical nodes of the cluster in the Remote Access
attribute of each virtual Client resource.
“Task 6: Configure privileges”
Set up NetWorker administrator privileges for each physical node and proxy
client in the cluster.
“Task 7: Configure a proxy client”
If a hardware provider such as the EMC VSS Provider is being used, a proxy
client must be configured for a clustered NMM client.
Note: There is an additional consideration for backing up a clustered Client resource: the proxy
client is not allowed to be a member of the cluster. This is by Microsoft design, it is not
supported by the Windows Cluster service. Microsoft describes this restriction in a MSDN
article at http://msdn.microsoft.com/en-us/library/aa384600(VS.85).aspx, in the subsection
“Auto-Import Hardware Shadow Copies Are Not Supported on Windows Cluster Service.”
Cluster failover and backups
operation will fail. The next scheduled backup operation will be the next valid
backup.
example, a CLARiiON storage solution may have an InActive snapshot as the result
of a failed backup. In this case, use the Navisphere user interface or the NaviCLI
command line interface to search for and delete the inactive snapshot.
298
Windows Server Cluster recovery
The Cluster recovery procedures are different for Windows Server 2003 and Windows
Server 2008. The following sections provide more information:
◆
“Windows Server 2003 clustered client disaster recovery” on page 299 provides
information about cluster recovery in Windows Server 2003.
◆
“Cluster recovery options” on page 91 describes the Cluster tab in NMM
Recovery options. This tab provides options for authoritative or nonauthoritative
restore. Authoritative restores of the Cluster Writer are only supported for
Windows Server 2008.
Windows Server 2003 clustered client disaster recovery
There are three different cluster recovery scenarios presented:
!
◆
“Disaster recovery of a cluster client” on page 299 describes how to recover a
cluster from a backup created by the NMM client.
◆
“Disaster recovery from a legacy VSS backup” on page 302 describes how to
recover a cluster from a VSS backup created before the NMM client was installed.
◆
“Disaster recovery of a clustered client from a legacy non-VSS backup” on
page 304 provides recommendations for how to handle legacy non-VSS backup
which does not work with the NMM client.
CAUTION
Ensure that your snapshot policies include at least one rollover to conventional
backup. Disaster recovery can only be performed from a conventional backup.
Chapter 14, “Snapshot Management,” describes how to roll over a snapshot to a
conventional backup.
Disaster recovery of a cluster client
The following procedure describes how to recover a single node in a two-node
cluster. The procedure assumes that you are recovering Node_B and that Node_A is
available.
To perform a disaster recovery of a cluster client:
1. Use the Cluster Administrator tool on Node_A to evict Node_B from the cluster.
For information about evicting a cluster, consult the Microsoft website or
documentation for the Cluster Administrator tool.
2. At the command prompt, type the following to clean up MSCS:
cluster node Node_B /forcecleanup
!
CAUTION
After cleaning the Cluster Server, you must reproduce the preexisting
application environment on Node_B so MSCS can administer the applications
as it did before the node failed. One way to reproduce the preexisting
application environment is to perform a recovery for each application that was
previously installed on the node.
Windows Server Cluster recovery
299
3. Use the Cluster Administrator tool to select Join an existing Cluster and type the
cluster name.
4. Use the Cluster Administrator tool on Node_A to ensure that Node_B appears
available for failover operations.
5. Stop the cluster service in Node_A by using one of the following methods:
• With the control panel service tool, stop the Cluster service.
• At the command prompt, type the following:
net stop clussvc
6. Select the required items to be recovered for Node_B. The procedure differs
depending on whether or not the cluster service resides on a domain controller.
If the cluster service resides on a member server and is not a domain controller:
a. Open the NMM client software on Node_B.
b. From the left pane, select Recover > System Recover Session.
c. Select all local drives except for the NMM client installation folder and the
Replication Manager (RM) installation folder.
d. Select the SYSTEM COMPONENTS folder.
e. From the System Recover Session toolbar, click Start Restore.
If the cluster service resides on a domain controller:
a. Boot the NMM client in directory service restore mode. “Booting Windows in
Directory Service Restore mode” on page 102 provides more information
about these recovery options.
b. Open the NMM client software.
click NetWorker.
!
CAUTION
300
g. Stop the cluster service in Node_B by using one of the following methods:
– With the control panel service tool, stop the Cluster service.
– At the command prompt, type the following:
net stop clussvc
h. Select all local drives except for the NMM client installation folder and the
Replication Manager (RM) installation folder.
i. In the navigation tree, select BootableSystemState and Cluster Writer.
j. Select the SYSTEM COMPONENTS folder.
k. From the System Recover Session toolbar, click Start Restore.
When prompted, restart Node_B.
7. Start the Cluster service on Node_A by using one of the following methods:
• With the control panel service tool, start the Cluster service.
net start clussvc
8. Use the Cluster Administrator on Node_A to confirm that the cluster resources
were recovered to the point in time when the backup occurred. If you are
performing regular scheduled backups, this will recover the cluster database to a
point shortly before the loss of Node_B.
9. Use the Cluster Administrator on Node_A to monitor the cluster "joining" status
of node_B.
10. Use the Cluster Administrator on Node_B to verify that the cluster group can be
moved between the nodes by right-clicking the group and selecting Move group.
Cluster recovery environment variable settings
If a Windows Server 2003 cluster database recover failed because the partition layout
of the quorum disk has changed since the backup, or the drive letter of the quorum
disk is already being used, the recovery process can be repeated with these
environment variables defined as:
◆
%NsrClusterForceRecover% should be set to 1.
◆
%NsrClusterNewQuorumDrive% should be set to the drive letter of the new
quorum disk (for example, R:).
There should only be one active node during the recovery. If other nodes are active
and the environment variable %NsrClusterForceRecover% is not defined, then the
recovery will fail. With %NsrClusterForceRecover% defined, the recover process will
attempt to shut down the cluster service on the other active nodes. It will also attempt
to change the signature and drive letter (if %NsrClusterNewQuorumDrive% is
defined) of the current quorum disk even if its partition layout does not match the
one described in the backup. However, the success of the recovery is not guaranteed
even with %NsrClusterForceRecover% defined.
301
Disaster recovery from a legacy VSS backup
The procedure for disaster recovery of a clustered client from a legacy VSS backup
differs depending on whether or not the cluster service resides on a domain
controller.
For cluster service on a member server
This recovery is for a VSS backup created before the NMM client was installed. This
backup would include VSS SYSTEM SERVICES, VSS SYSTEM FILESET, and VSS
SYSTEM BOOT.
The following procedure describes how to recover a single node in a two-node
cluster. The procedure assumes that you are recovering Node_B and that Node_A is
available:
1. Use the Cluster Administrator tool on Node_A to evict Node_B from the cluster.
!
CAUTION
After cleaning up the Cluster Server, you must reproduce the preexisting
application environment on Node_B so MSCS can administer the applications as it
did before the node failed. One way to reproduce the preexisting application
environment is to perform a recovery for each application that was previously
installed on the node.
3. Use the Cluster Administrator tool to select Join an existing Cluster and type the
cluster name.
• Using the control panel service tool, stop the Cluster service.
net stop clussvc
6. Select the required items to be recovered for Node_B:
a. Open the NMM client software on Node_B.
b. From the left pane, select Recover > System Recover Session.
c. Select all local drives except for the NMM client installation directory.
d. Select VSS SYSTEM SERVICES, VSS SYSTEM FILESET, and VSS SYSTEM
BOOT.
e. From the System Recover Session toolbar, click Start Restore.
f. When prompted, restart Node_B.
• Use the control panel service tool to start the Cluster service.
net start clussvc
302
8. Use the Cluster Administrator tool on Node_A to confirm that the cluster
resources were recovered to the point in time when the backup occurred. If you
have been performing regular scheduled backups, this will recover the cluster
database to a point in time shortly before the loss of Node_B.
9. Use the Cluster Administrator tool on Node_A to monitor the cluster "joining"
status of Node_B.
10. Use the Cluster Administrator tool on Node_B to verify that the cluster group can
be moved between the nodes by right-clicking the group and selecting Move
group.
For cluster service on a domain controller
This recovery is for a VSS backup created before the NetWorker VSS Client was
installed. This backup would include VSS SYSTEM BOOT, VSS SYSTEM FILESET,
and VSS SYSTEM SERVICE.
The following procedure describes how to recover a single node in a two node cluster.
The procedure assumes that you are recovering Node_B and that Node_A is
available:
1. Use the Cluster Administrator on Node_A to evict Node_B from the cluster.
!
CAUTION
After cleaning up the Cluster Server, you must reproduce the preexisting
application environment on Node_B so MSCS can administer the applications
as it did before the node failed. One way to reproduce the preexisting
application environment is to perform a recovery for each application that was
previously installed on the node.
3. Use the Cluster Administrator to select Join an existing Cluster and type the
cluster name.
• Use the control panel service tool to stop the Cluster service.
net stop clussvc
6. Select the required items to be recovered for Node_B:
a. Boot the NMM client in directory service restore mode. “Booting Windows in
Directory Service Restore mode” on page 102 provides more information
about these recovery options.
b. Open the NMM client software.
click the NetWorker tab.
303
!
CAUTION
g. Stop the cluster service in Node_B by using one of the following methods:
– Using the control panel service tool, stop the Cluster service.
– At the command prompt, type the following:
net stop clussvc
h. Select all local drives except for the NMM client installation directory.
i. Select VSS SYSTEM SERVICES, VSS SYSTEM FILESET, and VSS SYSTEM
BOOT.
j. From the System Recover Session toolbar, click Start Restore.
When prompted, restart Node_B.
• Use the control panel service tool to start the Cluster service.
net start clussvc
8. Use the Cluster Administrator tool on Node_A to confirm that the cluster
resources were recovered to the point in time when the backup occurred. If you
have been performing regular scheduled backups, this will recover the cluster
database to a point in time shortly before the loss of Node_B.
9. Use the Cluster Administrator tool on Node_A to monitor the cluster "joining"
status of Node_B.
10. Use the Cluster Administrator on Node_B to verify that the cluster group can be
moved between the nodes by right-clicking the group and selecting Move group.
Disaster recovery of a clustered client from a legacy non-VSS backup
This recovery is for a non-VSS backup created before the NMM client was installed.
This backup would include SYSTEM STATE, SYSTEM DB, and SYSTEM FILES.
Recovery of non-VSS backups of the cluster database by using the NMM client does
not work. To ensure that backups of the cluster database are available for recovery,
perform a full backup of the system directly after updating the system to the NMM
client. This creates a post-installation cluster backup that can be recovered by using
the NMM client.
304
Monitoring a
recovery
The progress of all recovery operations can be viewed from the Monitor window.
“Monitor view” on page 42 provides more information about the Monitor window.
305
306
13
Microsoft Active
Directory Backup and
Recovery
◆
◆
◆
◆
◆
◆
Active Directory backup and recovery ....................................................................
Performing an Active Directory or ADAM backup ...............................................
Performing a granular Active Directory or ADAM backup .................................
Recovering all Active Directory data .......................................................................
Performing a granular recovery of Active Directory objects and attributes ......
Active Directory disaster recovery ...........................................................................
308
309
310
318
320
324
307
Active Directory backup and recovery
(NMM) backup procedures in Chapter 2, “Configuring a Scheduled Backup,” and
recovery procedures in Chapter 4, “System Backup and Recovery,” with the specific
details needed to back up and recover Active Directory, Active Directory Application
Mode (ADAM), and Active Directory Lightweight Directory Services (AD LDS).
Note: In Windows Server 2008, AD LDS includes the ADAM functionality provided in
Windows Server 2003. Unless otherwise noted, assume that all references to ADAM in this
chapter also apply to AD LDS.
NMM supports the following types of backup:
◆
Active Directory Backup for Disaster Recovery:
• Backup of entire Active Directory database
• Utilizes Volume Shadow Copy (VSS) framework to create snapshot of Active
Directory data
• Backup Level: Full only
◆
Active Directory Granular Backup:
• Traditional/non-VSS backup of individual Active Directory objects and object
attributes
• Backup Levels: Full and Incremental
NMM supports the following types of recovery:
◆
Active Directory Recovery for Disaster Recovery:
• Recovery of entire Active Directory database
• Authoritative or Nonauthoritative Recovery Modes
◆
308
Active Directory Granular Recovery: Recovery of individual Active Directory
objects or attributes, including objects that are in the tombstone database.
Performing an Active Directory or ADAM backup
Performing an Active Directory backup enables the recovery of the entire Active
Directory database in the event of a catastrophic loss such as a hardware failure.
Active Directory data that is protected for a disaster is recoverable on an “all or
nothing” basis. You cannot choose individual objects for recovery, instead you must
recover the entire Active Directory database in one recovery operation.
To enable the recovery of Active Directory or ADAM data, create a snapshot-based
scheduled backup as described in Chapter 2, “Configuring a Scheduled Backup.”
When specifying save sets for the scheduled backup, follow these guidelines:
◆
Active Directory is backed up by normal system save operations. It is backed up
as part of SYSTEM COMPONENTS, as part of standard Windows Server
backups.
Note: Microsoft best practices also recommends that a backup Active Directory server be
set up so that it can be promoted to the primary Active Directory server, in the event of a
disaster.
◆
For an ADAM Client resource, specify the save set as:
APPLICATIONS:\ADAM <instance_name> Writer
where <instance_name> is the name of the ADAM instance.
Planning backups for disaster recovery
C:\
SYSTEM COMPONENTS:\
• Application software.
2. Back up the domain controller on the Active Directory machine.
Performing an Active Directory or ADAM backup
309
Performing a granular Active Directory or ADAM backup
A granular, or traditional non-VSS, Active Directory backup enables you to recover
individual objects and object attributes. A granular Active Directory Application
Mode (ADAM) backup enables you to recover individual application partitions.
WARNING
Granular recovery should only be performed as a supplement to the full Active
Directory database backup described in “Types of supported backup and recovery”
on page 308. A disaster recovery plan requires regular full Active Directory database
backups.
A granular backup does not use snapshot technology. Instead, the backup is routed
directly to a granular backup medium.
Considerations for granular backup
The following are general guidelines to consider before performing a granular Active
Directory backup:
310
◆
System-only attributes are not backed up with Active Directory objects, and
cannot be restored. System-only attribute values should not be changed.
◆
Changing the system date and time to an older date in the domain controller is
not recommended. Each item in Active Directory is marked by time. Active
Directory uses time to resolve any data conflicts. Restore of a deleted object with
the NMM client will fail if the date and time are changed after objects are backed
up. If a change in the system date or time is necessary, immediately take a backup
of the domain.
◆
The Group Policy object exists in two places: as an Active Directory object in the
Group Policy Container and in the Windows registry. Many of the settings that
can be edited from the Group Policy Management Console actually reside in the
system registry and not in Active Directory. Therefore, if these values are changed
after a backup, a recovery will not restore the attribute values and the system
values will take precedence.
◆
A restored user account is automatically set to disabled and the pwdLastSet
attribute is not recovered for security. A new password must be set after a user
account is restored.
◆
A change to a user object’s memberOf attribute is reflected for the target and not
the user’s properties. For example, if a user is added to the Guests group, the
Guests group object is modified, not the user object. If an incremental backup is
performed, the Guests object should be included in the backup and not the user
object.
◆
The rootDSE object is dynamic when created. ADAM lists the connection and the
primary registered partition. When a bind is requested on the rootDSE object,
ADAM returns the value of the default Naming Context. On a domain controller,
this is always the Domain Naming Context. Since the domain is visible in ADAM,
do not backup the domain using port 50000. Use the standard LDAP port 389.
◆
Do not modify the msNPAllowDialin attribute directly in Windows Server 2003,
SP1. Use the RAS administration function to modify the value. The Active
Directory Users and Computers (ADUC) MMC snap-in uses the RAS API to
modify this attribute and correctly modifies the userParameters attribute as well.
If the RAS API is not used, the userParameters attribute is not modified with the
msNPAllowDialin attribute and the configuration will be incomplete.
◆
Many configuration settings are stored in Active Directory, but LDAP cannot
always be used to write data directly. Also, some items stored in Active Directory
are only copies. Use the appropriate interfaces or APIs to modify settings.
Configuring Active Directory or ADAM granular backups
The tasks involved in configuring a granular Active Directory or ADAM backup are
similar to those for VSS writers supported by NMM. However, there are several tasks
that are not performed for granular Active Directory or ADAM backup.
Note: Do not configure a proxy client for granular ADAM or Active Directory backups.
To configure an Active Directory or ADAM granular backup, perform the following
tasks:
◆
“Task 1: Set up local Windows SYSTEM account as an ADAM administrator” on
page 311
This task is only required for granular ADAM backups. Skip this task for granular
Active Directory backups.
◆
“Task 2: Configure a pool for backup operations” on page 312
◆
◆
◆
◆
Task 1: Set up local Windows SYSTEM account as an ADAM administrator
The ADAM server, which is also a NetWorker client, relies on the NetWorker client’s
nsrexecd service to spawn the required binaries to perform a backup. The nsrexecd
service and other binaries run under the Windows SYSTEM account. To ensure that
these binaries can access ADAM data, the Windows SYSTEM account must be given
access rights to each instance on the ADAM server.
Note: These steps are not required for Active Directory backup.
To set up the Windows SYSTEM account with permissions on the local ADAM server:
1. Open the ADAM ADSI Edit utility.
The ADAM ADSI Edit utility is available with ADAM SP1.
2. Connect to the ADAM instance.
3. In the Connection Settings dialog box, select Well-known naming context and
select Configuration from the list box.
4. Type the port number (typically 50000) of the ADAM partition in the Port box and
click OK.
5. In the left pane, Select CN=Roles.
6. In the right pane, right-click on CN=Administrators and then select Properties.
311
7. In the CN=Administrators Properties dialog box, select the Member Attribute
from the attributes list, and then click Edit.
8. Click Add Windows Account and add the local ADAM Server name.
9. Save your changes.
Repeat this procedure for each ADAM instance that you want to back up.
Task 2: Configure a pool for backup operations
Create a backup pool to store the save sets that contain metadata.
Configuring a backup pool is a four-part process:
1. “Configuring a device” on page 312
2. “Configuring a label template” on page 313
3. “Configuring a backup pool” on page 313
4. “Labeling the device” on page 314
Note: All of the procedures in this section must be performed on a NetWorker server release
7.3.3 or later. Access the NetWorker server from the Administration window of the NetWorker
Management Console.
The EMC NetWorker Release 7.4 Service Pack 1 Multiplatform Version Administration
Guide provides more detailed information about performing the steps included here.
Configuring a device
Configure a media device for snapshot metadata. Configure a tape, file, or advanced
file type device. For best performance, configure a file or advanced file type device so
that data can be recovered without the potential delay associated with retrieving a
tape.
To configure a device:
1. In the NetWorker Administration window of the NetWorker Management
Console, click Devices on the taskbar.
2. In the expanded left pane, select Devices.
4. In the Name attribute, replace the default name with the path and name of the
device:
• If the device is configured on the NetWorker server’s storage node, the name
is the simple device path. For example, C:\tmp\d0 for a file type device.
A tape device would have a format similar to \\.\Tape0.
• If the device is configured on a remote storage node, the name must indicate
that the storage node is remote, by including rd= and the name of the remote
storage node in the device path. For example, if the remote storage node is
neptune, then the device path might be rd=neptune:c:\tmp\d0.
5. In the Comment field, specify a comment for the device.
6. From the Media Type attribute, select the appropriate media type.
7. Select the Auto Media Management attribute.
8. Click Configuration.
312
9. In the Target Sessions attribute, type or select a value. Set this attribute to a value
that will speed up the backup.
10. Click OK.
Configuring a label template
Labels identify the kind of data that is stored on the volumes in a backup pool. Label
templates define a naming convention for labels. Create a label template for volumes
that are used to contain snapshot metadata.
To configure a label template:
Media.
2. In the expanded left pane, select Label Templates.
4. In the Name attribute, type a name that identifies the series of labels as belonging
to volumes for metadata.
5. In the Comment attribute, specify a comment for the label.
6. In the Fields attribute, specify a text name such as Metadata and on a separate line,
specify a numeric range such as 001-999 or a text range such as aa-zz.
These attributes are used to incrementally identify each label.
7. Click OK.
Configuring a backup pool
Note: Use backup pools for granular Active Directory or ADAM backups that are different
than the pools used for NMM client backups that include snapshots. In this way, you can sort
granular Active Directory and ADAM backups from other backups.
To configure a backup pool to store snapshot metadata:
Media.
2. In the expanded left pane, select Media Pools.
4. For the Name attribute, type a name that matches the label template.
5. In the Comment attribute, specify a comment for the pool.
6. Select the Enabled attribute.
7. For the Pool type attribute, select the backup pool type.
8. For the Label template attribute, select the matching label template.
9. Do not select a group in the Groups attribute.
Do not associate the pool with a particular group. This pool may be associated
with several backup groups, so a single group should not be specified here.
10. Click the Selection Criteria tab.
11. For the Clients attribute, specify a NMM client. A value must be entered for this
attribute because no group was specified.
313
12. In the Devices attribute, select each device that can accept backups for this pool.
13. Click OK.
Labeling the device
Before a device can be used for backup, it must be labeled.
To label a device:
Devices.
2. In the right pane, right-click on the name of the device and select Label.
3. In the Pool attribute, select the backup pool that was created.
4. Select the Mount after Labeling attribute.
5. Click OK.
Task 3: Configure a backup schedule
Backup schedules determine the days on which full or incremental backups are run.
Full backups include all of the data specified in an NMM client resource save set,
while incremental backups include only the data that has changed since the last
backup.
Table 48 on page 314 contains the considerations for configuring a backup schedule
for an NMM client. The procedure for creating backup schedules for an NMM client
is the same as for creating a backup schedule for a regular NetWorker client.
Table 48
Considerations for NMM client backup schedules
Backup levels
Consideration
Full level
These backups are supported for all types of data.
Incremental
backups
These are supported only for backup of volume directories or paths
such as E:\ or E:\business_files\.
Note: Level 1 to 9 backups are not supported for granular Active Directory and ADAM
backups.
Backup groups enable you to balance backup loads to reduce the impact on your
storage and network resources. Use a backup group to assign the following attributes
to a set of Client resources:
◆
Backup start times
◆
Schedules
◆
Backup pools
Note: A Client resource or backup group cannot specify volumes that use both CLARiiON and
Symmetrix storage systems. For hosts that have both CLARiiON and Symmetrix volumes,
create separate backup groups and Client resources. Each backup group and its Client
resources can specify volumes from only one type of storage system.
314
To configure a backup group:
Configuration.
2. In the expanded left pane, select Groups.
4. In the Name attribute, type a name for the backup group.
5. In the Comment attribute, type a description.
6. For the Start Time attribute, enter the time when the first backup is to be created.
7. For the Autostart attribute, select Enabled.
Note: Unlike most other NMM application backup Client resources, snapshot policies are
not applicable to granular Active Directory or ADAM backups.
Do not specify or assign a snapshot policy or a snapshot pool to the backup
group:
• Make sure that the Snapshot attribute is clear, so that there is no check mark
beside it.
• Do not select a snapshot policy in the Snapshot Policy attribute.
• Do not select a pool in the Snapshot Pool attribute.
8. Click the Advanced tab.
9. Set the Client Retries attribute to 0 (zero).
10. Click OK to create the backup group.
A Client resource specifies what to include in a backup of an NMM client. Client
resources are associated with other backup resources such as groups.
For Active Directory and ADAM granular backup, there are some differences from
other NMM backup configurations.
Note: Do not use the same Client resource for Active Directory or ADAM granular backups
and other NMM client backups including Active Directory or ADAM backups for the purposes
of disaster recovery.
You can create multiple Client resources for the same NMM client host. In this way,
you can apply different backup attributes to different types of information on the
same host. For example, if the NMM client host is an Exchange Server, you can create
one Client resource to back up the Exchange databases and create another Client
resource to back up Windows system component data. This way, you can back up
Exchange databases many times a day, and back up Windows system component
data once a day.
To create a Client resource for granular Active Directory or ADAM backup:
Configuration.
315
4. In the Name attribute, type the fully qualified hostname of the NetWorker client
computer.
9. In the Save Set attribute, specify the components to be backed up. :
Specify domain objects in the following format:
DC=corp, DC=domain,DC=com
For example:
DC=corp, DC=xyz, DC=com
where the backup saves the entire domain named corp.xyz.com from its root
level.
will be added.
13. In the Backup command attribute, type the appropriate command for Active
Directory objects or ADAM objects:
• Active Directory objects
Type the following command in the Backup Command attribute:
nsradsave.exe
Active Directory domain objects and application partitions on the domain host
cannot be backed up in the same Client resource. Each type of data must be
specified on a separate Client resource. For example, to back up an Active
Directory domain or part of a domain, type the following in one Client
resource:
nsradsave.exe
• ADAM objects
To back up ADAM application partitions, type the following command in the
Backup Command attribute, specifying the port number 50000. For example,
type the following:
nsradsave.exe -p 50000
316
14. If the NMM client is part of a cluster, type the names of the physical nodes of the
cluster in the Remote Access attribute.
16. In the Aliases attribute, validate that the NETBIOS name for the client is present.
This will be automatically populated by NetWorker when name resolution is
configured. If the NETBIOS name is not present, add the NETBIOS name for the
client.
17. Click OK.
Task 6: Configure privileges
In addition to the permission set up in “Task 1: Set up local Windows SYSTEM
account as an ADAM administrator” on page 311 for performing granular ADAM
backup, in both ADAM and Active Directory granular backup the NMM client must
be granted NetWorker Administrator privileges in order to perform media database
operations during a deletion. If you are setting up a NMM client in a cluster, grant
NetWorker administrator privileges to each cluster node and virtual server.
To add configure privileges:
Configuration.
2. In the expanded left pane, click User Groups.
3. Right-click the Administrators group, then select Properties. The Properties
dialog box appears.
4. In the Users attribute, add the following values for the NMM client host. Place
each value on a separate line:
user=administrator,host=NMM_client_host
user=system,host=NMM_client_host
where NMM_client_host is the DNS hostname of the NMM client.
If the NMM client is installed in a cluster, grant NetWorker Administrator
privileges to each cluster node and virtual server.
5. Click OK.
317
Recovering all Active Directory data
The steps in this procedure assume that any damaged hardware has been replaced
and that all software has been reinstalled.
To ensure the successful recovery of the Active Directory server, the backup steps
described in “Performing an Active Directory or ADAM backup” on page 309 must
have been completed.
If a disaster recovery is necessary, “Active Directory disaster recovery” on page 324
provides information.
Note: If you have followed Microsoft best practices, you will have a backup Active Directory
server that can be promoted to the primary Active Directory server in the event of a disaster. If
you do not have a backup Active Directory server that can be promoted to the primary Active
Directory server, complete the steps in this section to recover from a disaster of the Active
Directory server.
To recover Active Directory server data:
1. Before you can recover Active Directory server data, you must boot the NMM
client in Directory Service Recovery mode:
• If you have not booted the NMM client in Directory Service Recovery mode,
proceed to step 2 on page 318.
• If you have already booted the NMM client in Directory Service mode,
proceed to step 7 on page 318.
2. On the NetWorker client domain controller, close all programs and restart
Windows.
When the computer restarts, a list of startup choices appears. The startup choices
and the duration of the startup display are based on the settings in the operating
systems section and the boot loader section of the boot.ini file.
For Windows Server 2003 this is boot.ini. For Windows Server 2008 it is called the
Boot Configuration Data (BCD) Store.
3. Select the Windows boot option for the domain controller.
4. Press F8 to display a list of special boot options.
5. Select Directory Service Restore Mode (Windows Domain Controllers only)
from the list of special boot options.
When you boot in this mode, Active Directory is taken offline.
6. Log in as administrator. Use the password that was specified when the domain
controller was created.
Windows starts in safe mode.
7. Open the NMM client software and select the NetWorker Server on which the
NMM client was configured for backup.
8. In the left pane, select Recover > System Recover Session.
9. From the navigation tree, open the SYSTEM COMPONENTS folder.
backup.
318
To recover from a previous backup:
• To view all versions of a backup object prior to the selected browse time,
right-click an object in the navigation tree and select Versions.
10. Select items to recover:
• If the Active Directory server is not in a clustered environment, select all of the
items under the SYSTEM COMPONENTS folder.
• If the Active Directory server is part of a cluster environment:
a. From the Options menu, select System Recover Session Options and then
click the NetWorker tab.
b. Clear the attribute titled Use Microsoft best practices for selecting the
system state. Click Yes when asked to confirm your selection.
c. Click OK to close the System Recover Session Options dialog box.
d. Select all of the items under the SYSTEM COMPONENTS folder except for
the Cluster Writer.
11. To determine whether any volumes need to be mounted, right-click the selected
item and select Required Volumes.
13. To perform an authoritative recovery of the Active Directory server, complete the
procedure for authoritative recovery of NTDS or FRS data. “Performing an
authoritative recovery of NTDS or FRS data” on page 103 provides more
information about these types of recovery.
14. After the Active Directory recovery is complete, reboot the host to restart the
Active Directory server.
15. If the Active Directory server is part of a cluster environment, select the Cluster
Writer for recovery.
16. Recover any remaining volume or application data. Chapter 4, “System Backup
and Recovery,” provides more information about recoveries.
Recovering all Active Directory data
319
Performing a granular recovery of Active Directory objects and attributes
You can individually select deleted Active Directory objects and their attributes for
recovery.
Note: Due to Active Directory limitations, there are some restrictions on the recovery of objects
and attributes. “Recovery Restrictions for Active Directory” on page 321 provides more
information about recovery restrictions.
To recover an Active Directory object or object attribute:
1. Start the NMM client program.
2. From the application toolbar, click the NetWorker server icon to select the
3. In the left pane, select Recover > Active Directory Recover Session.
4. In the navigation tree, select the Active Directory objects to be recovered. By
default, the objects displayed in the navigation tree are from the most recent
backup.
To search for an item, click the Search tab. “Searching for an item” on page 44
provides details about searching for a recovery item.
5. To determine whether any volumes need to be mounted for a selected object,
right-click an object and select Required Volumes.
6. Recover the entire object or selected object attributes:
• To recover the entire object:
a. Select the object.
b. From the Active Directory Recover Session toolbar, click Start Recover to
begin the recovery operation.
• To recover just selected attributes of an object:
a. Right-click an object and select Recover this Item.
The Active Directory Recover Attributes dialog box appears.
b. Select each attribute that is to be recovered.
If all attributes or no attributes are selected, the entire object is recovered.
c. Click OK.
320
Recovery Restrictions for Active Directory
The following restrictions apply when recovering Active Directory objects and
attributes:
◆
“Tombstone lifetime restriction” on page 321
◆
“System-only attributes cannot be recovered” on page 321
◆
“Object password attributes” on page 322
◆
“Attributes with null values” on page 323
◆
“Schema objects” on page 323
Note: After starting an Active Directory recovery, access the Monitor view to verify the status
of the recovery.
The Microsoft website provides more information about Active Directory objects and
attributes.
Tombstone lifetime restriction
When an Active Directory object is deleted, it is sent to a special container, named
Deleted Objects. This object is now invisible to normal directory operations, and is
called tombstone. It is better to recover a tombstone object instead of re-creating it
because data such as the Security Identifier (SID) and the Global Unique Identifier
(GUID) are stored with the tombstone object. This data is critical for additional data
recoveries, such as assigned group permissions. For example, Access Control Lists
(ACLs) use the SID of a security identifier object to store its permissions. A re-created
group would get a new SID and GUID so that permissions assigned to the old group
would be lost. Similarly, the SID and GUID are both used to recover a user profile. A
user's profile would become unusable if a user with the same name is re-created. This
is because the new profile would be given a new SID and GUID.
Objects in tombstone are deleted when they reach the tombstone lifetime age for the
domain, which is, by default, 180 days. After an object is deleted from the tombstone,
it cannot be recovered. This is an Active Directory restriction. The tombstone lifetime
is a configurable attribute of a Windows domain.
System-only attributes cannot be recovered
Object attributes that are system-only can neither be backed up nor recovered. This is
an Active Directory restriction. Table 49 on page 321 provides a sample list of
system-only attributes are not backed up.
Table 49
System-only attributes that are not backed up
badPwdCount
lastLogon
uSNChanged
badPasswordTime
logonCount
uSNCreated
distinguishedName
objectCategory
userAccountControl
dSCorePropagationData
objectClass
whenChanged
instanceType
objectGUID
whenCreated
lastLogoff
sAMAccountType
321
Attributes that are retained for a deleted object
Table 50 on page 322 provides a sample list of attributes are retained for an Active
Directory object when it is deleted and moved to the tombstone database.
Table 50
Retained attributes after object is deleted
attributeID
mSMQOwnerID
subClassOf
attributeSyntax
name
systemFlags
distinguishedName
nCName
trustAttributes
dNReferenceUpdate
objectClass
trustDirection
flatName
objectGUID
trustPartner
governsID
objectSid
trustType
groupType
oMSyntax
userAccountControl
instanceType
proxiedObjectName
uSNChanged
IDAPDisplayName
replPropertyMetaData
uSNCreated
legacyExchangeDN
sAMAccountName
whenCreated
mS-DS-CreatorSID
securityIdentifier
These attributes are restored when deleted objects from the tombstone database are
restored. Objects that do not retain all of their mandatory attributes would cause a
constraint violation error during a restore attempt. For example, a published shared
printer has mandatory attributes (printerName, serverName, shortServerName,
uNCName, and versionNumber) which are not retained in the tombstone database.
Note: To recover objects that typically have mandatory attributes deleted, set the search flags of
the specific attributes to eight. When the object is deleted, the attributes with search flag set to
eight are transferred to the tombstone database. The entire object can then be restored from
tombstone with the NMM client.
Object password attributes
An object’s password is not recovered. After recovering an object with a password
attribute, the Windows administrator must reset the password.
Moved or renamed objects
If an object is moved to another location or renamed, but not deleted from Active
Directory they cannot be restored even if they are successfully backed up. This is due
to the objects not being stored in the deleted storage database (tombstone). Only
objects in that database can be restored.
However, the attributes for these objects can be restored from the Context menu of
the Active Directory interface, which restores the objects with the specific attribute
sets.
322
Attributes with null values
Attributes with null values are not backed up and therefore, are not recovered. For
example, if an attribute for a phone number was empty (null), then the null Phone
Number attribute would not be backed up.
This is an Active Directory restriction and is intended to prevent the unintentional
overwriting of valid attribute values. For example, if a Phone Number attribute was
null when a snapshot was taken but later a valid phone number was added,
subsequent recovery operations would not overwrite the valid phone number with a
null value.
Schema objects
Schema objects cannot be recovered and therefore, should never be deleted.
323
Active Directory disaster recovery
A disaster recovery would require a complete system recovery in addition to an
application data restore. Performing a disaster recovery requires that proper
measures were taken during backup of Active Directory data. “Planning backups for
disaster recovery” on page 309 provides information for backing up data for disaster
recovery.
The disaster recovery procedure for Active Directory is generally a system recovery
followed by a standard Active Directory recovery. The following sections provide
information:
324
◆
Chapter 4, “System Backup and Recovery,” provides system component recovery
information.
◆
“Recovering all Active Directory data” on page 318 and “Performing a granular
recovery of Active Directory objects and attributes” on page 320 provide recovery
information.
14
Snapshot
Management
◆
◆
◆
Roll over a snapshot.................................................................................................... 326
Roll back a snapshot ................................................................................................... 327
Delete a snapshot ........................................................................................................ 330
Snapshot Management
325
Snapshot Management
Roll over a snapshot
A rollover is the process of creating a conventional backup from a snapshot.
Conventional backups include backup to tape, file type, or advanced file type
devices. Rollover policies are typically controlled automatically through snapshot
policy settings on the NetWorker server. Chapter 2, “Configuring a Scheduled
Backup,” provides more information about scheduling backups.
However, you can perform a rollover manually if necessary. For example, before
recovering a database that currently exists only as a persistent snapshot, consider
rolling over the snapshot to tape to protect the data in case the snapshot becomes
corrupted during the recovery process.
Another reason to perform a rollover manually, is to roll over only selected save sets
from the snapshot. For example, if a snapshot contains two save sets: D:\ and E:\,
you can select to roll over only the data in the E:\ save set.
To roll over a snapshot:
NetWorker client was configured for backup.
2. In the left pane, select Snapshot Management.
3. In the navigation tree, select the snapshot and expand the snapshot to view the
save sets. Select the snapshot or snapshot save sets to be rolled over.
4. From the Snapshot Management toolbar, click Rollover to begin the rollover
operation.
5. In the left pane, select Monitor to view the progress of the rollover operation.
326
Snapshot Management
Roll back a snapshot
A rollback recovery is the process of returning data to an earlier point in time. The
entire contents of a persistent snapshot is recovered during a rollback recovery. You
cannot select individual components in a rollback recovery operation. A rollback
recovery is the fastest way to recover the contents of a persistent snapshot.
A rollback recovery is known as a "destructive" recovery because all changes that
were made after the selected snapshot was taken are overwritten. For example,
suppose that a snapshot of a database is taken every hour. At 5:30 P.M., it is
determined that the database is corrupt and that the last known consistent database
snapshot was taken at 5:00 P.M. When the database host is rolled back to the 5:00 P.M.
snapshot, all database transactions between the time of the 5:00 P.M. snapshot and
5:30 P.M. are overwritten.
Rollback recovery requirements
To perform a rollback operation, the following requirements must be met:
◆
The snapshot to be rolled back must reside on a CLARiiON or Symmetrix storage
system that is supported with the EMC VSS Provider. EMC NetWorker Module for
Microsoft Applications Release 2.1 Installation Guide provides more information.
◆
The NMM client to which the snapshot is to be rolled back must be supported
with a proxy client and the snapshot must have been created using a proxy client.
Chapter 2, “Configuring a Scheduled Backup,” provides more information about
backup using a proxy client.
◆
The nested mount points on file systems are dismounted prior to performing the
rollback restore.
◆
The Microsoft Exchange checkbox titled “This database can be overwritten by a
restore” is selected for each Exchange database that is to be recovered. This
checkbox can be found by using the Exchange SYSTEM Manager application.
◆
The databases are detached for each Microsoft SQL Server database that is to be
recovered.
◆
NMM does not support rollback restore for backups performed with EqualLogic
or Celerra hardware. Only conventional restore and snapshot restore are
supported for those hardware.
◆
Distributed Link Tracking Client must be turned off during rollback, or else
dismount will fail. Turn off the Distributed Link Tracking Client in Windows
Services before starting rollback.
◆
The SYSTEM process must not be running during rollback, or else dismount will
fail. The SYSTEM process leaves the volume open when there are software
shadow copies on the volume. To ensure that the volumes are not open, delete
any shadow copies on the volumes that are being rolled back.
Storage array descriptor file
A storage array descriptor file, sa.ini, contains information about the storage arrays
connected to the proxy client data mover. The sa.ini file is created with the
nsr_sa_ini.exe utility, and placed in the nsr\res directory on the NMM client host.
Nsr_sa_ini.exe prompts you for information about the storage arrays connected to
the proxy client.
327
Snapshot Management
The required inputs are different for CLARiiON and Symmetrix:
◆
For a CLARiiON storage array, you must provide an array user name and
password. When you complete the entries for a CLARiiON sa.ini file, the
password string is encrypted when you save the file.
“Sample CLARiiON sa.ini file” on page 328 provides an example of the entries in
this file.
◆
For a Symmetrix storage array, the sa.ini file has fewer parameters. You do not
need an account name, password, or IP address.
“Sample Symmetrix sa.ini file” on page 329 provides an example of the entries in
this file.
To create a storage array descriptor file:
1. At the command line, type:
nsr_sa_ini.exe
2. Follow the prompts to provide the array information.
Note: Even though the utility appears to allow you to enter credentials for more than one
storage array, only the first one will be used by NMM. You must create a separate sa.ini for
each storage array.
When you have completed entering the array information, the file is saved to the
nsr\res directory on the NMM client host.
!
IMPORTANT
Do not edit the sa.ini file with a text editor to make changes in the array
parameters. Using a text editor can corrupt the file, or compromise the password
entries. Use the nsr_sa_ini.exe tool to edit the sa.ini file if you need to make
changes after you create it.
Sample CLARiiON sa.ini file
A sample sa.ini file for a CLARiiON storage array would look similar to the
following:
[Storage_Array]
Storage_Array_Count=2
Storage_Array_1=APM00052106710
[APM00052106710]
Array_Type=CLARiiON
Array_model=CX500
Array_user=sbarma
Array_pass=****
Array_pass_confirm=****
Array_SPA=122.222.111.128
Array_SPB=122.222.111.129
328
Snapshot Management
Sample Symmetrix sa.ini file
A sample sa.ini file for a Symmetrix storage array would look similar to the
following:
[Storage_Array]
Storage_Array_Count=1
Storage_Array_1=000187461273
[000187461273]
Array_type=Symmetrix
Array_sid=197
Array_Model=DMX-800
System rollback may fail with CLARiiON if the configuration information is not current
If a CLARiiON array is migrated, or the information in the CLARiiON configuration
is not up to date, rollback may fail. If there is more than one entry in the CLARiiON
clarcnfg file, Replication Manager only uses the first entry. If the clarcnfg file is not
current, the first entry may be pointing to a CLARiiON that is no longer connected to
the system.
Edit the clarcnfg file so that there is only one entry, pointing to the current CLARiiON
you want to use.
Performing a rollback recovery
Before performing a rollback recovery, ensure that the rollback requirements have
been met. “Rollback recovery requirements” on page 327 provides more information
about these requirements.
To perform a rollback recovery:
1. Open the NMM client and select the NetWorker server on which the NetWorker
client was configured for backup.
3. If you are performing a rollback for EMC Symmetrix configurations that use
RAID5 Business Continuance Volumes (BCVs), perform the following steps:
a. Delete all snapshots for the Standard EMC Disk (STD), except for the snapshot
to be rolled back. “Delete a snapshot” on page 330 provides information about
deleting snapshots.
The deletion operation will put the snapshot BCVs in the established state.
Once you have completed the deletions, there should be only one snapshot
BCV in the split state, and that this is the snapshot BCV that is to be rolled
back.
Note: To prevent the loss of data from a snapshot that is to be deleted, perform a
rollover of the snapshot before the deletion operation. “Roll over a snapshot” on
page 326 provides more information about rolling over a snapshot.
b. Use the SYMCLI commands to disassociate the BCVs and Virtual Devices
(VDEVs) that are in the established state.
c. If the snapshot used multiple STDs, perform the previous steps for each STD.
4. In the navigation tree, select the snapshot to be rolled back.
329
Snapshot Management
5. Right-click the selected snapshot and select Rollback.
6. In the left pane, select Monitor to view the progress of the rollback operation.
7. When rollback processing has completed, reboot the host to complete the rollback
operation.
8. If you performed a rollback on an Exchange or SQL database, remount the
database. If you performed a rollback on a file system that had nested mount
points, remount the nested mount points.
Delete a snapshot
Snapshot deletion policies are typically controlled automatically through snapshot
policy settings on the NetWorker server. Chapter 2, “Configuring a Scheduled Backup,”
provides more information about policy settings.
However, you can delete a snapshot manually. For instance, if space must be made
available immediately on the storage system, you can manually roll over and then
delete snapshots.
!
CAUTION
If a snapshot is deleted without the use of the NMM client user interface, the
snapshot may still appear to be recoverable in the NMM client user interface. If
such a snapshot is selected for recovery, the recover operation will fail. A snapshot
could by in used by a user in another application. A snapshot could also be deleted
by VSS or the storage subsystem if resources are exhausted due to high
input/output rates on the volumes.
To delete a snapshot:
NetWorker client was configured for backup. “Opening the software and
connecting to a NetWorker server” on page 43 provides more information about
opening NMM and connecting to a NetWorker server.
3. In the navigation tree, select the snapshot to be deleted.
4. Right-click the selected snapshot and select Delete.
5. In the left pane, select Monitor to view the progress of the delete operation.
330
15
NetWorker Client
Management
◆
◆
Specifying ranges for service and connection ports............................................... 332
Specifying a password................................................................................................ 333
331
Specifying ranges for service and connection ports
A service port is a listener port that provides services to NMM client hosts. The
default range for service ports is 7937 to 9936. A connection port is used to contact a
service, whether it is on a NetWorker server, storage node, or client. The default range
for connection ports is 10001–30000.
To implement an enhanced security environment, it may be necessary to reduce the
range of available ports. EMC NetWorker Multiplatform Version Administration Guide,
release 7.3.3 or later, provides more information about determining the size of the port
range.
To reduce the range of available ports:
1. From the Options menu, select Configure Options.
The Configure Options dialog box appears.
2. In the Service Ports attribute, enter the range of available ports.
3. In the Connection Ports attribute, enter the range of available ports.
4. Click OK.
Note: If a firewall exists between the NetWorker client and any NetWorker servers, ensure that
the firewall is configured to accept the port ranges entered in the Configure Options dialog
box.
332
Specifying a password
The password is used to protect items that will be backed up with either PW1
password-protection, or PW2 encryption and password-protection. PW1 and PW2
protection is enabled on the NetWorker server with a local or global directive. EMC
NetWorker Multiplatform Version Administration Guide, release 7.3.3 or later provides
more information about PW1 and PW2 protection.
!
CAUTION
Changing the password does not change the password for files that have already
been backed up. To recover PW1 or PW2 password-protected files, you must
provide the password that was in effect at the time of the backup.
To change the PW1 password:
1. From the Options menu, select Configure Options.
The Configure Options dialog box appears.
2. In the Old password attribute, type the password that is currently in effect.
3. In the New password attribute, type the new password.
4. In the Confirm password attribute, retype the password.
5. Click OK.
The new password will be applied to future scheduled backups for the NMM client
where password-protection has been enabled with a global or local directive.
Specifying a password
333
334
16
Invisible Body Tag
Troubleshooting
◆
◆
◆
◆
◆
◆
The troubleshooting process......................................................................................
NMM client error messages.......................................................................................
NMM client issues.......................................................................................................
Checking log files ........................................................................................................
Manually stopping and starting services.................................................................
Other troubleshooting resources...............................................................................
Troubleshooting
336
338
340
347
350
351
335
Troubleshooting
The troubleshooting process
NetWorker and the NMM provide many resources to help discover and isolate
problems. Setting up these resources before problems occur can help if and when
troubleshooting problems occur later.
If the problem is not with the NMM client, there are other NetWorker resources and
documentation that may provide a solution.
The following flowchart, Figure 60 on page 336, provides an overview of the process
and sections of this chapter for troubleshooting a scheduled backup or recovery
problem.
Find error
NetWorker Server
NetWorker
Client-side or
Server-side?
NetWorker Client
No
Unique to
NMM Client
(volumes and
writers)?
Yes
Check other EMC
documentation
Does the error
appear in a log file
or message
window?
No
Yes
Check NMM Client
error messages
section
Fixed?
Check NMM
Client issues
section
Yes
Finish
No
Try a
different
solution?
No
Contact
Technical
Support
Yes
GEN-000753
Figure 60
336
Troubleshooting process
Troubleshooting
Setting up notifications
Set up NetWorker software and NMM client software to send notifications about
events. These notifications can be sent to several locations, including log files, SNMP
printers, SNMP management console, and email messages. The EMC NetWorker
Multiplatform Version Administration Guide, release 7.3.3 or later, describes how to set
up notifications.
Finding errors
When a problem occurs, a notification or message might be displayed, or appear in a
log file. Table 51 on page 337 describes several ways that NetWorker server, NMM
client, and other NetWorker features provide information about problems.
Table 51
Notifications and errors
Type
Description
Email notification of a
NetWorker event
NetWorker client can be set up to send an email notification about NetWorker
events.
Error message windows
Appear on the NetWorker server when an error occurs during backup or recovery
operations.
NetWorker Monitor window
Displays details of current NetWorker server activities, operations related to devices
and libraries, and events that require user intervention.
Event viewer
Alerts users that user intervention is required on the NetWorker server.
Log files
Log files are created for each operation and process. Separate log files are
available for NetWorker and various features, depending on what features are
installed and in use:
• NetWorker Module for Microsoft Applications
• NetWorker
• NetWorker PowerSnap
• Replication Manager
• Solutions Enabler
• EMC VSS Provider
The logging level of the log file is controlled by the debug level of the process.
The troubleshooting process
337
Troubleshooting
NMM client error messages
Error messages help to identify which product is having a problem. The following
sections describe error messages that are specific to the NMM client.
E_VOLUME_NOT_SUPPORTED
This error occurs when CLARiiON is down, an array is unmanaged, or the
CLARiiON disk is not visible. This error also occurs when Data Mover is used to take
a local volume replica. The error message will appear in NMM client and Replication
Manager logs.
Solution
Check in the CLARiiON array for storage allocation.
E_VETO_PROVIDER
There are several possible causes for this error:
◆
Storage resources such as Symmetrix clones or CLARiiON BCVs have been added
but the Replication Manager service was not recycled to rebuild the storage
resource cache. The provider finds the LUN as hosted by the storage array, but
fails to find storage resources in the storage resource cache.
◆
There is a load on the CLARiiON array, and replica creation fails.
◆
A CLARiiON snapshot is performed when there are not enough LUNs in the
reserved LUN pool. In this case, the rmsnap command fails.
◆
A CLARiiON snapshot is performed, but the number of established snapshot
sessions to the LUN has already reached the limit.
Solution
Use the net stop and net start commands to stop and restart the EMC VSS provider,
Microsoft VSS services, and Replication Manager services.
For example, to stop and restart the Replication Manager service:
1. From the command line, stop the rmagentps service:
net stop rmagentps
2. Start the rmagentps service:
net start rmagentps
VSS_E_WRITERERROR_RETRYABLE error code 0x800423f3
This error occurs if a savegroup is rerun, and it was stopped previously while a
replica was being taken and the replica did not complete.
Solution
Stop EMC VSS provider, Microsoft VSS services, and Replication Manager services.
Then start those services. Also restart Exchange Information store if it was running
and was backed up.
338
Troubleshooting
VSS CLIENT... Invalid Writer Selection... for APPLICATIONS
This error might occur if Microsoft Exchange services are not up or if SQL databases
are offline.
Another cause of this error is if there is a typographical error in writing
APPLICATIONS:\Microsoft Exchange Writer.
Solution
For Microsoft Exchange applications, start the Exchange services. If Exchange
services were already up, dismount and mount the databases and then start the
savegroup.
For SQL Server applications, bring the databases online.
Solution
Retype the words correctly or use the nsrsnap_vss_save -? command, as described in
“Displaying valid application data save sets” on page 63, to display a listing of all
valid application save sets on the NMM client.
access to the path is denied
This error might occur if enough space is not available for staging in Temp directory
during SharePoint granular recovery.
Solution
The staging area required for both SharePoint granular backup and recovery must be
the size of the largest site-collection, regardless of what is being backed up or
recovered. For example, when there is a 500 Gb site-collection and a 50 Gb sub-site,
and only the sub-site is being recovered, 500 Gb of staging space is required. Thus,
the staging space requirements are determined by the actual requirements for a
backup or recovery.
Object reference not set to an instance of an object
This error might occur when during SharePoint granular backup, the nsr_moss_save
–f command is used.
Solution
Make sure that the NetWorker SharePoint Connector Service has sufficient rights to
SharePoint farm.
NMM client error messages
339
Troubleshooting
NMM client issues
The following section describes some common problems and possible solutions for
NMM client. Check this section first when there appears to be a problem but there is
no specific error message in monitoring or event logs.
Savegroup failed in scheduled backup
A notification appears in the NetWorker Monitor window that a savegroup failed.
Solution
Check the savegroup details for the failed save set. This may provide an exact cause,
or a general error, which should indicate whether it is a client-side or server-side
issue. The EMC NetWorker Multiplatform Version Administration Guide, release 7.3.2 or
later, provides information about viewing group backup details.
If there is not enough information in the savegroup details, check the NMM client log
and other client-side logs.
“Checking log files” on page 347 provides more information about the log names,
locations, and details.
In progress backup fails
operation will not be successful. The next scheduled backup operation will be the
next valid backup.
Solution
example, a CLARiiON storage solution may have an inactive snapshot as the result of
a failed backup. In this case, use the EMC NaviCLI interface to search for and delete
the inactive snapshot.
NetWorker is blocked by Windows Firewall
After installing Microsoft Windows Server 2003 (SP1), the Windows Firewall may be
blocking NetWorker processes from sending information through the firewall.
Solution
provides information about how to extend the Security Configuration Wizard or
manually configure Windows Firewall exceptions.
Need to run utility to obtain Exchange Server information
The utility nwexinfo.exe runs when NMM client is installed on an Exchange Server.
This utility gathers the Exchange username, password, and domain, and then inserts
this information in the registry.
340
Troubleshooting
Solution
Rerun the utility if any of the following conditions occurs:
◆
Exchange Server is installed after NMM client is installed.
◆
The Exchange username, password, or domain values change after the NMM
client is installed.
To run the utility:
1. Navigate to the following folder: <NetWorker_install_path>\nsr\bin
2. Run nwexinfo.exe.
3. Change the Exchange password, username, and domain as required.
Requesting a snapshot for a volume on hardware that is
not transportable
When volumes are added to the NMM client snapshot set, an error might appear if a
transportable hardware snapshot is requested and the volume that is being replicated
is not a volume that resides on special hardware that is transportable. The following
error message might also appear:
VSS_E_VOLUME_NOT_SUPPORTED_BY_PROVIDER.
Solution
The Exchange server and proxy server (if one is set up for Exchange) must have
matching volumes or alternate mount path in PowerSnap (PS) attribute. An Exchange
backup that is mounted on a proxy server needs either matching volumes or an
alternate location to mount Exchange databases and logs.
An example of matching volumes would be D: to D: and E: to E:.
An example of alternate location would be "G:\altpath" on the proxy server. In this
example, the snapshot of "D:\" would be mounted on the proxy server as
"G:\altpath\D\" and the snapshot of "E:\" would be mounted on the Proxy server as
"G:\altpath\E\."
To use the alternate mount path, set the PowerSnap attribute NSR_ALT_PATH in the
Client resource. In the preceding example of an alternate location, the attribute would
be set to "NSR_ALT_PATH=G:\altpath." “Task 5: Configure a Client resource” on
page 61 provides more information about creating a Client resource.
PowerSnap service must be restarted after the NetWorker service
is restarted
After restarting the NetWorker service, if the PowerSnap service is not restarted, the
first retention or rollover will fail. The following message appears:
RPC send operation failed. A network connection could not be
established with the host.
The PowerSnap service will re-establish the connection on the next operation, so this
failure will only occur once.
Solution
To fix this problem, restart the PowerSnap service after restarting the NetWorker
service, before running a PowerSnap operation.
NMM client issues
341
Troubleshooting
PowerSnap validation for recovery fails
When the service is not stopped, recovery fails while validating PowerSnap for
recovery. Some VSS writers require that the service be stopped while recovering. For
example, the SQL writer requires that the service be stopped while recovering the
SQL database.
Solution
To fix this problem, stop the service before recovering.
Snapshot backup may fail with writer errors
A snapshot backup may fail with VSS writer errors that indicate the writer may be in
an unstable state, such as the following:
VSS_E_WRITERERROR_NONRETRYABLE
VSS_E_BADSTATE
VSS_E_UNEXPECTED_WRITER_ERROR
Solution
Contact EMC Technical Support.
Microsoft I/O write error when taking a replica in a CLARiiON array
This problem might occur when a replica is taken in a CLARiiON array LUN backup.
Solution
To fix this problem, stop the EMC VSS provider, Volume Shadow Copy Service and
Replication Manager services. Then start those services.
Creating a replica takes a long time in CLARiiON environment
Older version of CLARiiON FLARE® OS that is running on CLARiiON may take a
long time to create a replica.
Solution
To fix this problem, update CLARiiON FLARE OS to the latest version.
Cannot recover Exchange database due to overwrite prevention error
If an Exchange database has not been selected for recovery in the Exchange System
Manager, recovery for that database will fail. When a recovery attempt fails, several
entries are displayed in the NetWorker Monitor window. For example:
"VSS CLIENT... Database "Mailbox Store 2 (SERVERX)" in Storage
Group "Second Storage Group" does not have the "This database
can be overwritten by a restore." checkbox marked."
"VSS validatation and preprocessing failed for Microsoft
Exchange Writer."
"nsrsnap_vss_recover: Recovery operation failed. Refer to log
file(s) for details."
342
Troubleshooting
The following message may be displayed by Microsoft Exchange:
The database files in this store were replaced with older
versions by an offline restore. To use the restored files, open
the database property page for this store, select "this
database can be overwritten by a restore", wait for active
directory replication, and try again.
Error ID: C104173A
Exchange System Manager
Solution
Search for "KB293324"on the Microsoft Help and Support website, and go to the
article "XADM: "C104173A" Error Message Occurs After You Restore the Database."
Follow the instructions for changing the setting in Exchange System Manager.
Incremental backups within NetWorker Module for Exchange
Client may be promoted to full backups
If the NMM client is installed, requests for an incremental backup through the
NetWorker Module for Exchange Client are automatically promoted to a full backup.
This does not affect NMM client incremental backups of volumes.
For example, suppose that NetWorker Module for Exchange Client requests an
incremental backup for Exchange server "EXSRV1." If NetWorker NMM client is
installed, then the NetWorker Module for Exchange Client promotes its own
incremental backup to a full backup to maintain the integrity of the backup and
prevent data loss. However, the NetWorker NMM client can still perform an
incremental backup of a volume, such as "D:\" on "EXSRV1."
Incremental backups within NetWorker Module for SQL Server
Client may be promoted to full backups
If the NetWorker NMM client is active and configured to back up the SQL database,
requests for an incremental backup through the NetWorker Module for SQL Client
are automatically promoted to a full backup to maintain the integrity of the backup
and prevent data loss. Installation of the NMM client itself does not automatically
promote NetWorker Module for SQL Client incremental backups to full backups.
This does not affect NMM client incremental backups of volumes.
For example, suppose that the NetWorker Module for SQL Client requests an
incremental backup for SQL server "SQLSRV1." If the NMM client is installed, and
the NetWorker NMM client performs backups of the SQL database, then the
NetWorker Module for SQL Client automatically promotes the incremental backup to
a full backup.
However, if the NMM client is installed but does not perform backups of the SQL
database, then a request through the NetWorker Module for SQL Client for an
incremental backup is allowed. In either case, the NMM client can still perform an
incremental backup of a volume, such as "D:\" on "SQLSRV1."
NMM client issues
343
Troubleshooting
Diagnosing VSS Writer Issues
NMM provides a command to dump the Microsoft VSS Writer metadata to XML files.
These XML files detail what each active writer on the system is doing, and may be
useful to support personnel for investigating and analyzing writer issues.
To dump the VSS Writer metadata to XML files, run the nsrsnap_vss_save command
with the -G command switch, for example:
D:\Legato\nsr\bin>nsrsnap_vss_save -G
When you run the command a message displays the directory the files were sent to.
For example:
NMM .. requesting of writer metadata completed successfully and
deposited in C:\Users\ADMINI~1\AppData\Local\Temp\3\xmlfiles\
The directory contains an .XML for each writer on that system, for example:
Exchange Writer.XML
SQL Writer.XML
etc.
The files will probably be hidden because the directory they are sent to are hidden by
default. You can make the folder visible by using Start > Control Panel > Folder
Options > View > Show Hidden Files and Folders.
Double-click an XML file to view it in your browser, such as Internet Explorer or
Firefox.
You can also copy the files to another folder. For example:
D:\>mkdir xmlfiles
D:\>cd xmlfiles
D:\xmlfiles>copy
C:\Users\ADMINI~1\AppData\Local\Temp\3\xmlfiles\*.*
C:\Users\ADMINI~1\AppData\Local\Temp\3\xmlfiles\ExchangeWriter
.xml
C:\Users\ADMINI~1\AppData\Local\Temp\3\xmlfiles\SQL Writer.xml
2 file(s) copied.
Modifying the Exchange 2007 COM+ component’s user account or password
If your user account, your password information, or both change after the COM+
component is registered, you can use the following procedure to change the user
account or password information for the COM+ component:
1. Start component services.
2. Expand Computers > My Computer > DCOM Configuration.
3. Right-click Replication Manager Exchange Interface and select Properties from
the context menu.
344
Troubleshooting
4. Click the Identity tab.
5. Modify the domain user account or password and click OK.
This will also modify the account information for the Replication Manager Exchange
Interface Service.
Save sets may not appear in the recover window
NMM client uses the host machine NETBIOS or "short" name when connecting to the
NetWorker server to browse backups. If the NETBIOS name is not found, NMM will
Solution
Add the NETBIOS name to the Aliases attribute of the Client resource. “Task 5:
Configure a Client resource” on page 61 provides more information about creating a
Client resource.
Recovery of AES-encrypted files fails without passphrase
If an application such as SQL Server or Exchange Server is backed up with AES
encryption, and then a recovery is attempted without the correct passphrase,
recovery will fail. The application files will only be partially recovered, and will be
empty or in a corrupted state.
Solution
During data recovery, you must specify the pass phrase used at the time of backup if
it is not the default or current pass phrase. Provide the passphrase when recovering
AES encrypted files, using NMM Security recovery options. “Security recovery
options” on page 100 provides more information about specifying pass phrases in
NMM.
Unable to restore files encrypted with both Microsoft Windows Encrypting File System and AES
encryption
When AES encryption is applied to a file that is also encrypted using the Microsoft
Windows Encrypting File System (EFS), the backup will be reported as successful.
However, recovery of the file will fail and the following message will be written to
the NetWorker log file:
recover: Error recovering <filename>. The RPC call completed
before all pipes were processed.
Solution
Do not use AES encryption when backing up files that are encrypted using EFS.
NMM client issues
345
Troubleshooting
Anti-virus programs block recovery
When a recovery from a save set of “all” is attempted, anti-virus programs may block
NMM recovery. Anti-virus programs are designed to protect their own program files
and settings from external threats that may be attempting to disable the program. The
anti-virus program cannot always distinguish the recovery from an attack.
Solution
Prior to recovery, disable the anti-virus programs protection properties. The name
and location of the settings varies, but look for a setting like “Prevent modification of
<Anti-Virus Program Name> command agent files and settings.” Clear or disable the
setting. Check the anti-virus program’s documentation or online help for information
about finding and disabling this setting.
To recover files to a system protected by an anti-virus program:
1. Disable the anti-virus program’s setting for protecting its own files and settings.
2. Perform the recovery.
3. Reset the anti-virus program to protect its files and settings.
346
Troubleshooting
Checking log files
There are many log files generated by NMM client and associated features. These log
files include errors that occur during the processes. Check the log files in the
following order:
1. “NMM client” on page 347
2. “PowerSnap client” on page 347
3. “Replication Manager” on page 348
4. “Solutions Enabler” on page 348
5. “Active Directory” on page 349
6. “NetWorker server” on page 349
In addition, third party providers generate their own logs in place of the Solutions
Enabler log (hwprov.log). For more information, check the third party
documentation.
NMM client
Log file
The NMM client log file is named: nmm.raw
Location
Applogs folder. For example, C:\Program Files\Legato\nsr\applogs\.
Description
Both save and recover CLIs write to this file. The log file is cumulative, so it is
appended with each run. The logging level of the log file is controlled by the debug
level of the process (-D).
PowerSnap client
Log files
NetWorker VSS Client Interface Library: libvsspsclnt.XXXXX.log
PowerSnap Client Core Processes. There are log files for each process that runs:
◆
nsrbragent.XXXXX.log
◆
nsrsnapagent.log
◆
nsrsnapbwragent.log
◆
nsrsnapck.log
Location
Logs folder. For example, C:\Program Files\Legato\nsr\logs\.
Description
Usually the PowerSnap Client creates one set of log files for each backup or recover
session, and a debug and trace log are created. These log files are tied together by
using the Snapshot Session ID.
Checking log files
347
Troubleshooting
The logging level of the log files is controlled by setting a variable in
NSR_PS_DEBUG_LEVEL in NetWorker Server, in client configuration.
Error messages in PowerSnap Client log files also appear in the standard user
interface outputs for NMM client.
Replication Manager
Log files
The Replication Manager log files are:
◆
erm_clientXXXXX_debug.log
◆
erm_clientXXXXX_detail.log
◆
erm_clientXXXXX_summary.log
Location
Replication Manager Agent Service log files are in the logs\client folder of the RM
installation. For example, C:\Program Files\EMC\rmagentps\logs\client\.
Description
The log files are cumulative, and will wrap once they reach a certain length.
The debug level can be set through the registry key
HKLM\SOFTWARE\RMSERVICE\RMAgentPS\Client\4.0.0\CC_DEBUG_LEVEL.
The default value is 2, but can be changed to 3. To change the value, stop RM Services,
change the value in the registry, and then restart RM Services.
Error messages in Replication Manager log files also appear in the standard user
Solutions Enabler
Log files
The log files include a log file for the EMC VSS Provider and SYMAPI:
◆
hwprov.log
◆
symapi-<date>.log
Location
The Solutions Enabler product creates log files in the SYMAPI\logs folder of the
Solutions Enabler installation. For example, C:\Program Files\EMC\SYMAPI\log\.
Description
The log files are cumulative, or based on the date. The logging level of the EMC VSS
Provider log file is controlled by a registry key.
Error messages in Solutions Enabler log files do not appear in the standard user
348
Troubleshooting
Active Directory
Log files
The log files associated with Active Directory are:
◆
nsradsave.log
◆
nsradrecover.log
Location
Applogs folder. For example, C:\Program Files\Legato\nsr\applogs\.
NetWorker server
Log files
The NetWorker server creates several log files, which are documented in the EMC
NetWorker Multiplatform Version Administration Guide, release 7.3.2 or later.
Checking log files
349
Troubleshooting
Manually stopping and starting services
NMM client processes may need to be manually stopped and started in conjunction
with configuration and troubleshooting activities. The NetWorker server creates
several log files. These log files are documented in the EMC NetWorker Multiplatform
Version Administration Guide, release 7.3.2 or later, which provides more information
about starting and stopping services on Windows and UNIX hosts, Console server,
NetWorker Server, Client, or Storage Node.
Table 52 on page 350 lists the services for NetWorker, PowerSnap, and Replication
Manager that are active on the system.
Table 52
Executable
Product
Category
Function
nsrexecd
NetWorker
Service
Authenticates and processes the NetWorker server’s remote
execution requests and executes the save programs on the
client.
nsrpsd
PowerSnap
Service
Provides PowerSnap Client services, including snapshot
consistency check function, to apply retention policy and backup
functions to rollover snapshots.
irccd.exe
Replication
Manager
Service
Provides Replication Manager Client application and storage
services for creating VSS-based snapshots and for rollback of
VSS-based snapshots for EMC storage.
RM_ExchangeInterface.
exe
Replication
Manager
Service
Provides Replication Manager Server Exchange Interchange
services to handle Exchange Server 2007-specific commands
for Replication Manager.
rm_api.dll
Replication
Manager
Library
Library that provides Replication Manager Client interface library
for NetWorker, used by NetWorker backup and recover
processes.
nsrsnap
PowerSnap
Process
Temporary process active during snapshot groups that provides
client-side workflow for snapshot groups, including applying
retention policy, spawning process to create the snapshot, and
spawning process to rollover the snapshot.
nsrsnap_vss_save
NetWorker
Process
Temporary process active during snapshot groups that is
responsible for executing the client-side workflow to create the
snapshot for the save sets specified in the Client resource. It is
spawned by ‘nsrsnap’ and will communicate with irccd.exe to
create the VSS-based snapshot, and then will communicate with
PowerSnap services to register the snapshot.
nsrcscd.exe
NetWorker
Service
This service acts as an agent to provide details about a system
environment to the NMM user interface.
Libpsvssclnt.dll
PowerSnap
Library
Library that provides PowerSnap Client Interface for NetWorker
backup and recover processes.
nsrsnapagent
Temporary
Service
Temporary service active during snapshot backup and restore
operations that provides snapshot management functions to
import and assign drive letters to a snapshot.
nsrsnapbwragent
Temporary
Service
Temporary service active during snapshot backup and file
system snapshot browse operations that provides the ability to
read the file system on the snapshot.
Process
Provides graphical user interface for NetWorker software.
winworkr
350
Services used in NMM client
NetWorker
Troubleshooting
Other troubleshooting resources
Does the problem appear to be on the NetWorker server or on a NetWorker client?
If the problem appears on the NetWorker server, or otherwise seems to be on the
server side, check to make sure that the NetWorker server is installed and configured
correctly. Also, check the log files and error message documentation for the
NetWorker server.
If the problem appears to be with PowerSnap or Replication Manager, and is not
unique to NMM client, check other documentation. The following documents
provide specific error message, troubleshooting, or other documentation outside of
NMM client:
◆
EMC NetWorker Multiplatform Version Installation Guide, release 7.3.2 or later
◆
EMC NetWorker Multiplatform Version Administration Guide, release 7.3.2 or later
◆
EMC NetWorker Multiplatform Version Error Message Guide, release 7.3.2 or later
◆
EMC NetWorker PowerSnap Module Installation and Administration Guide
◆
EMC Replication Manager Administrator’s Guide
Other troubleshooting resources
351
Troubleshooting
352
Glossary
This glossary contains terms related to disk storage subsystems. Many of these terms
are used in this manual.
A
ad hoc backup
administrators group
See manual backup.
Microsoft Windows user group whose members have the rights and privileges of
users in other groups, plus the ability to create and manage the users and groups in
the domain.
ASM (application
specific module)
Program that is used in a directive to specify how a set of files or directories is to be
backed up or recovered. For example, compressasm is a NetWorker directive used to
compress files.
archive
Backing up directories or files to an archive volume to free disk space. Archived data
is not recyclable.
archive volume
auto media
management
autochanger
autochanger sharing
Volume used to store archive data. Archived data cannot be stored on a backup
volume or a clone volume.
Feature that enables the storage device to automatically label, mount, and overwrite
an unlabeled or recyclable volume.
See library.
See library sharing.
B
backup
Operation that saves data to a volume.
See also conventional backup and snapshot.
backup components
metadata document
See metadata document.
backup group
See group.
backup level
See level.
backup volume
Volume used to store backup data. Backup data cannot be stored on an archive
volume or a clone volume. See also volume.
353
Glossary
bootstrap
browse policy
Save set that is essential for NetWorker disaster recovery procedures. The bootstrap
consists of three components that reside on the NetWorker server: the media
database, the resource database, and a server index.
NetWorker policy that specifies how long backed-up data will be readily available for
recovery. Backed-up data that has not exceeded its browse policy time can be
recovered more quickly than data that has exceeded its browse policy time but not its
retention policy time. See also retention policy.
C
carousel
client
client file index
client-initiated
backup
See library.
Computer, workstation, or fileserver whose data can be backed up and recovered.
Database that tracks every database object, file, or file system that is backed up. The
NetWorker server maintains a single client index file for each client.
See manual backup.
Client resource
NetWorker Server resource that identifies the save sets to be backed up on a client.
The Client resource also specifies information about the backup, such as the schedule,
browse policy, and retention policy for the save sets. See also client and resource.
clone
Reliable copy of backed up data. Unlike volumes created with a simple copy
command, clone volumes can be used in exactly the same way as the original backup
volume. Single save sets or entire volumes can be cloned.
clone volume
cluster
Exact duplicate of a backup volume. One of four types of volumes that NetWorker
software can track (backup, archive, backup clone, and archive clone). Save sets of
these different types may not be intermixed on one volume.
1. Two or more independent network servers that operate and appear to clients as if
they are a single unit. The cluster configuration enables work to be shifted from one
server to another, providing "high availability" that allows application services to
continue despite most hardware or software failures. Also known as an agent (Sun),
logical server (HP TruCluster), package (HP-UX), and virtual server (Microsoft).
2. Group of disk sectors. The operating system assigns a unique number to each
cluster and keeps track of files according to which clusters they use.
command line
component
Line on a display screen, also known as a command prompt or shell prompt, where
you type software commands.
1. Group of related data that must be treated as a single unit for backup and recovery.
2. In Microsoft VSS terminology, a component is a subordinate unit of a writer.
components
metadata document
consistent
Console server
354
See metadata document.
State of a dataset that is fully and immediately available to an application view.
Software program that is used to manage NetWorker servers and clients. The Console
server also provides reporting and monitoring capabilities for all NetWorker
processes.
Glossary
conventional backup
See nonpersistent snapshot.
D
domain controller
data mover (DM)
data retention policy
datawheel
datazone
Dynamic Drive
Sharing (DDS)
device
Computer that stores directory data and manages user interactions within a domain,
including logon, authentication, directory searches, and access to shared resources.
Client system or application, such as NetWorker, that moves the data during a
backup, recovery, or snapshot operation. See also proxy client.
See retention policy.
See library.
Group of hosts administered by a NetWorker server.
Feature that allows NetWorker software to recognize shared drives.
1. Storage unit that reads from and writes to backup volumes. A storage unit can be a
tape device, optical drive, autochanger, or file connected to the server or storage
node.
2. When dynamic drive sharing (DDS) is enabled, refers to the access path to the
physical drive.
DFS (Distributed File
System)
Microsoft Windows add-on that allows you to create a logical directory of shared
directories that span multiple machines across a network.
directed recovery
Method of recovery that recovers data that originated on one client computer and
re-creates it on another client computer.
directive
disk subsystem
Instruction that directs NetWorker software to take special actions on a given set of
files for a specified client during a backup or recovery operation. Directives are
ignored in manual (unscheduled) backups.
Integrated collection of storage controllers and/or HBAs, disks, and any required
control software that provides storage services to one or more hosts, such as
CLARiiON arrays.
F
file index
file system
See client file index.
1. The software interface used to save, retrieve, and manage files on storage media by
providing directory structures, data transfer methods, and file association.
2. The entire set of all files.
full backup
See level.
G
group
Client or group of client computers that are configured to back up files at a
designated time of day.
355
Glossary
H
high-available system
host ID
System of multiple computers configured as cluster nodes on a network that ensures
the application services continue despite a hardware or software failure. Each cluster
node has its own IP address with private resources or disks that are available only to
that computer.
Serial number that uniquely identifies a host computer.
I
inactivity timeout
incremental backup
instant backup
instant restore
Number of minutes to wait before a client is considered to be unavailable for backup.
Backup level in which only files that have changed since the last backup are backed
up. See also level.
Process of creating a point-in-time copy (snapshot) of data from a single client and
saving it on a primary storage volume, which can be immediately recovered as a
backup copy.
Process of copying data created during an instant backup to its original location, or to
an alternate location, during a recover operation.
J
jukebox
label
legacy method
level
library
library sharing
local cluster client
locale settings
LUN (logical unit)
356
See library.
Electronic header on a volume used for identification by NetWorker or other data
mover application.
Use of special-case Microsoft APIs to back up and recover operating system
components, services, and applications.
Backup configuration option that specifies how much data is saved during a
scheduled or manual backup. A full (f) backup backs up all files, regardless of
whether they have changed. Levels one through nine [1-9] backup files that have
changed since the last lower numbered backup level. An incremental (incr) backup
backs up only files that have changed since the last backup.
Hardware device containing one or more removable media drives, as well as slots for
pieces of media, media access ports, and a robotic mechanism for moving pieces of
media between these components. Libraries automate media loading and mounting
functions during backup and recovery. The term library is synonymous with
autochanger, autoloader, carousel, datawheel, jukebox, and near-line storage.
Shared access of servers and storage nodes to the individual tape drives within a
library.
NetWorker client that is not bound to a physical machine, but is instead managed by
a cluster manager. It is also referred to as a logical or virtual client.
Settings that specify the input and output formats for date and time, based on local
language conventions.
Logical unit of storage on a CLARiiON system. This refers to a device or set of
devices, usually in a CLARiiON storage array.
Glossary
LUN address
SCSI identifier of a logical unit number (LUN) within a device target. Each LUN
address identifies a device on a SCSI bus that can perform input/output (I/O)
operations.
M
manual backup
Backup that a user performs from the client, also known as an unscheduled backup or
an ad hoc backup. The user specifies the files, file systems, and directories to back up.
media
Physical storage medium, such as magnetic tape, optical disk, or file system to which
backup data is written.
media database
media index
Database that contains indexed entries of storage volume location and the life cycle
status of all data and volumes managed by the NetWorker server. See also volume.
See media database.
metadata document
VSS Information stored in an XML document that is passed from the writer to the
requestor. Metadata includes the writer name, files and components to back up, a list
of components to exclude from the backup, and the methods to use for recovery. See
also shadow copy set.
mount
To make a database available for use or to place a removable tape or disk volume into
a drive for reading or writing.
mount point
See volume mount point.
N
NDMP (Network Data
Management
Protocol)
NetWorker
administrator
NetWorker client
TCP/IP-based protocol that specifies how heterogeneous network components
communicate for the purposes of backup and recovery.
User who can add to or change the configuration of the NetWorker server, media
devices, and libraries. NetWorker administrators must have their usernames
included in the NetWorker server Administrator list.
See client.
NetWorker Console
server
See Console server.
NetWorker
Management
Console
See Console server.
NetWorker server
NetWorker storage
node
nonclone pool
Computer on a network running the NetWorker software, containing the online
indexes, and providing backup and recover services to the clients on the same
network.
See storage node.
Pools that contain data that has not been cloned.
357
Glossary
nonpersistent
snapshot
Snapshot backup that is moved to secondary storage on the NetWorker server or
storage node and is no longer available for instant restore from a supported type of
primary storage.
O
online indexes
operator
Databases located on the NetWorker server that contain all the information
pertaining to the client backups (client file index) and backup volumes (media
database).
Person who monitors the server status, loads backup volumes into storage devices,
and executes day-to-day NetWorker tasks.
P
pathname
persistent snapshot
Set of instructions to the operating system for accessing a file. An absolute pathname
indicates how to find a file starting from the root directory. A relative pathname
indicates how to find the file starting from the current directory.
Snapshot that is retained on disk. A persistent snapshot may or may not be rolled
over to tape.
PiT (point-in-time
copy)
Fully usable copy of a defined collection of data, such as a consistent filesystem,
database, or volume, which contains an image of the data as it appeared at a single
point in time. A PiT copy is also called a shadow copy or a snapshot.
policy
Set of constraints that specify how long the save sets for a client are available for
recovery. Each client has a browse policy and a retention policy. When the retention
policy expires, the save sets associated with that policy are marked recyclable.
pool
PowerSnap
provider
proxy client
Feature to sort backup data to selected volumes.
EMC technology that provides point-in-time snapshots of data to be backed up.
Applications that are running on the host system continue to write data during the
snapshot operation, and data from open files is included in the snapshots.
Software component defined by Microsoft VSS, that plugs into the VSS environment.
A provider, usually produced by a hardware vendor, enables a storage device to
create and manage snapshots.
Surrogate client that performs the NetWorker save operation for the client that
requests the backup. A proxy client is required to perform a serverless backup.
Q
quiescing
Process in which all writes to disk are stopped and the file system cache is flushed.
Quiescing the database prior to creating the snapshot provides a transactionally
consistent image that can be remounted without file system checks or database
consistency checks. Quiescing a database is the most common way of creating a
database snapshot.
R
recover
358
To recover files from a backup volume to a client disk.
Glossary
Registry
requestor
replica
resource
Microsoft Windows database that centralizes all Windows settings and provides
security and control over system, security, and user account settings.
Interface with the Microsoft VSS infrastructure to initiate the creation and destruction
of shadow copy. NetWorker software is a requestor.
See shadow copy.
Component that describes the NetWorker server or its clients. Clients, devices,
schedules, groups, and policies are all NetWorker resources. Each resource has
attributes that define its properties.
restore
Process of retrieving individual datafiles from backup storage and copying the files to
disk.
retention policy
NetWorker policy that specifies the minimum period of time that must elapse before
backed-up data is eligible to be overwritten on the backup media. Backed-up data
that has not exceeded its browse policy time can be recovered more quickly than data
that has exceeded its browse policy time but not its retention policy time. See also
browse policy.
retrieve
rollback restore
rollover
root
To locate and recover archived files and directories.
Process by which a specific point-in-time copy (snapshot) of data is restored to the
source location by using the hardware's particular capabilities. A rollback restore is a
destructive save set restore.
Process of backing up a snapshot to a conventional backup medium such as tape.
Whether or not the snapshot is retained on disk depends on the snapshot policy.
Highest level of the system directory structure.
S
save set
SSID (save set ID)
save set recover
save set status
server index
serverless backup
Group of files or a file system from a single client computer, which is backed up on
storage media.
Internal identification number assigned to a save set.
To recover data by specifying save sets rather than by browsing and selecting files or
directories.
NetWorker attribute that indicates whether a save set is browsable, recoverable, or
recyclable. The save set status also indicates whether the save set was successfully
backed up.
See client file index.
Backup method that uses a proxy client to move the data from primary storage on the
application server host to secondary storage on another host. Serverless backups free
up resources on the application server by offloading the work of processing
snapshots to a secondary host.
service port
Port used to listen for backup and recover requests from clients through a firewall.
shadow copy
Temporary, point-in-time copy of a volume created using VSS technology. See also
Volume Shadow Copy Service (VSS).
359
Glossary
shadow copy set
shadow copy
technology
skip
Complete roadmap of what was backed up at a single instant in time. The shadow
copy set contains information about the writers, their components, metadata, and the
volumes. A backup components metadata document containing that information is
created and returned to the requestor after the snapshot is complete. Networker uses
this document with the corresponding save set at recover time.
Defined and standard coordination between business application, file system, and
backup application that allows a consistent copy of application and volume data to
exist for replication purposes.
Backup level in which designated files are not backed up.
snap clone
Exact copy of a snap set data backup. The clone operation is an archive operation
without the deletion of the source data. A new snap ID is assigned to the cloned copy.
snap ID
Also known as a snapid, a unique 64-bit internal identification number for a snap set.
snap set
Group of files, volumes, or filesystems from a single client, describing the collection
of data for which a point-in-time copy is created on an external disk subsystem, such
as a storage array.
snapshot
snapshot expiration
policy
snapshot policy
Point in time, read-only copy of data created during an instant backup.
Policy that determines how long snapshots are retained before their storage space is
made available for the creation of a new snapshot.
Set of rules that control the lifecycle of a snap set. The snapshot policy specifies the
frequency of snapshots, and how long snapshots are retained before recycling.
snapshot retention
policy
Policy that determines how many PIT copies are retained in the media database and
thus are recoverable.
staging
Moving data from one storage medium to a less-costly medium, and later removing
the data from its original location.
stand-alone device
Storage device that contains a single drive for backing up data. Stand-alone devices
cannot store or automatically load backup volumes.
storage device
storage node
See device.
Storage device physically attached to a computer other than the NetWorker server,
whose backup operations are administered from the controlling NetWorker server.
V
volume
1. A unit of physical storage medium, such as a magnetic tape, optical disk, or file
system to which backup data is written.
2. An identifiable unit of data storage that may reside on one or more host disks.
volume ID
volume mount point
volume name
360
Internal identification that NetWorker software assigns to a backup volume.
Disk volume that is grafted into the namespace of a host disk volume. This allows
multiple disk volumes to be linked into a single directory tree, and a single disk or
partition to be linked to more than one directory tree.
Name assigned to a backup volume when it is labeled. See also label.
Glossary
volume pool
Volume Shadow Copy
Service (VSS)
VSS
VSS component
See pool.
Microsoft technology that creates a point-in-time shadow copy of a disk volume.
NetWorker software backs up data from the shadow copy. This allows applications
to continue to write data during the backup operation, and ensures that open files are
not omitted.
See Volume Shadow Copy Service (VSS).
Subordinate unit of a writer.
W
writer
Database, system service, or application code that provides metadata document
information about what to back up and how to handle VSS components and
applications during backup and recovery operations. A Writer provides information
to requestors to ensure that application data is consistent, application files are closed
and ready for a slight pause to make a Shadow Copy.
361
Glossary
362
Index
A
Active Directory (AD)
backups for disaster recovery 309
backups for object selection granularity 310
disaster backup and recovery 38
granular backup and recovery 38
log files 349
Active Directory Lightweight Directory Services (AD LDS)
93, 308
active node 222
AD LDS backup and recovery 308
AD recovery
all objects for a disaster recovery 318
granular recovery 320
individual objects and attributes 320
null value restriction 323
restrictions for 321
schema object restriction 323
system-only attribute restriction 321
tombstone restriction 321
ADAM
authoritative restore 94
nonauthoritative restore 94
recovery options 93
AES encryption
conflict with EFS encryption 100
All save set
considerations for 84
alternate location
SharePoint directed recovery 183
alternate mount path for proxy client 200
alternate storage group, Exchange recovery to 235
application data save sets
considerations for 51, 122
considerations for SharePoint 2003 161
application information attribute
applications 65
Cluster Writer 297
DPM 254
Exchange 200
Hyper-V 283
SharePoint 2003 SQL data 156
SharePoint 2007 139
SQL 118
Application Specific Module (ASM) pass phrase 100
application writers
DPM 251
Exchange 197
Hyper-V 278, 281
SharePoint 2003 153
SharePoint 2007 136
supported 31
authoritative restore
ADAM 94
Cluster Writer 299
DFS 95, 96
FRS 97
NTDS 99
Avamar 73
configuration 75
B
backup
Active Directory 310
LCR 195
SharePoint 2003 152
backup and recovery
Hyper-V parent partition 273
backup client resource
configuring 61, 315
configuring DPM 256
configuring Hyper-V 286
configuring multiple 61, 256, 286, 315
backup command attribute
applications 85
DPM 257
Exchange 210, 211, 213, 214, 216
Hyper-V 287
SharePoint 2007 144, 146, 147
SharePoint 2007 granular 176
backup groups
configuring 59, 314
backup levels,
deduplication 76
backup pool
creating for snapshots 52, 312
363
Index
backup roadmap
Hyper-V 274
backup save sets
applications 61
DPM 252
Exchange 198
Hyper-V 282, 286
SharePoint 2007 137
SharePoint 2007 disaster recovery 170
specifying 84, 208, 211, 213, 214, 216, 256, 316
SQL 115
backup schedule
configuring 58, 314
backup time
selecting for recovery 45
backups
deduplication 76
best practices
SQL backup and recovery 121
best practices for backup and recovery
Exchange 203
Hyper-V child partition 279
SharePoint 2003 161
SQL in SharePoint 2003 161
browse time
C
child partition
best practices for backup and recovery 279
Hyper-V configurations 274
child partitions 273
CLARiiON
E_VETO_PROVIDER error message 338
error message 338
provider overview 32
replica takes long time 342
client resource
configuring for backup 61, 315
configuring for deduplication 78
configuring for DPM backup 256
configuring for Exchange backup 206
configuring for Hyper-V backup 286
configuring for SharePoint 2003 backup 157
configuring for SharePoint 2007 backup 143
configuring for SQL backup 119
configuring when in a cluster 69, 298
Cluster Continuous Replication (CCR)
Exchange backup 195
cluster server
entering physical hostname of 85, 257, 317
scheduling backups for 69, 298
Cluster tab 299
Cluster Writer recovery
authoritative restore 299
for Windows Server 2008 91
364
options 91
common provider
See EMC VSS Provider 29
component system writers
supported 31
configuration database
backup 135
SharePoint 2007 129
SharePoint 2007 database backup 147
SharePoint 2007 recovery 149
connection port range
specifying 332
connection to NetWorker server 43
consistency checker
utility 68
conventional recovery
overview 36
conventional snapshot policy example 57
D
data deduplication 72
deduplication
Avamar integration 73
backup configuration notes 77
backups 76
configuration 75
configuring client 78
prerequisites 73
querying deduplication save sets with mminfo 80
support 74
unsupported 74
Deduplication Node
configuration 75
deduplication, data 72
deleting snapshots 330
device
configuring for rollover backup 52, 312
DFS recovery options 94
directed recovery
DPM-protected server 264
Exchange Server 2007 238
NMM Client 105
SharePoint 2007 restore to an alternate location 183
system 105
directory services restore mode
booting in for FRS recovery 102
booting in for NTDS recovery 102
disaster recovery
backup 37
clustered client from a legacy non-VSS backup 304
DPM 267
Hyper-V 292
legacy VSS backup 302
Microsoft Exchange Server 2007 241
NMM client 109
NMM Client not on a domain controller 110
NMM Client on a domain recovery 111
planning for active directory 309
recovery for active directory 318
Index
SharePoint 2007 170
SQL 123
support in NMM 38
SYSTEM COMPONENTS 109
Windows Server 2003 clustered client 299
DNS resolution, NetWorker Deduplication Node 75
DPM
configuring client resource for backup 256
directed recovery 264
disaster recovery options 269
URL encoding for save sets 253
individual objects and attributes 267
E
EFS encryption
conflict with AES encryption 100
email notification
NetWorker events 337
EMC VSS Provider
overview 29
encryption
recovering backup data 86
errors
E_VETO_PROVIDER 338
E_VOLUME_NOT_SUPPORTED 338
error code 0x800423f3 338
Microsoft I/O write 342
NMM client 338
replica in a CLARiiON array 342
VSS CLIENT... Invalid Writer Selection 339
VSS_E_VOLUME_NOT_SUPPORTED_BY_PROVIDE
R 341
VSS_E_WRITERERROR_RETRYABLE 338
with Exchange 339
Event viewer 337
events
notification 340
exceptions
Windows Firewall 340
Exchange
client resource configuration 206
consistency checker utility 68
errors 339
point-in-time recovery 223
proxy client 201
recover to Recovery Storage Group (RSG) 226
roll-forward recovery 222
Exchange backup
CCR environment 195
LCR environment 195
Exchange recovery
alternate storage group 235
database can be overwritten selection 219
options for 219
original storage group 227
Remote Storage Group (RSG) 226
Exchange Server 2007
exclusion representation of save sets 40
F
failover from the passive node 222
file system
recovery options 89, 91
FRS recovery
booting in directory services restore mode 102
consideration for 87
options 97
full backup level
consideration for 58, 314
G
granular backup
SharePoint 2007 175, 176
granular backup and recovery
support in NMM 38
granular recovery
backup 37
DPM objects 260
SharePoint 2007 177
groups
configuring for backup 59, 314
guest. See also child partition
H
hardware providers 32
host
Hyper-V 274
Host. See also Hyper-V parent partition
Hyper-V
backing up a virtual machine with multiple volumes
284
backup roadmap 274
components used for backup and recovery 281
configuring backups 285
configuring client resource for backup 286
disaster recovery 292
granular backup and recovery 38, 280
host 274
parent partition disaster recovery 292
recovery 288
save sets 281
VSS writer 281
Hyper-V best practices
child partition backup and recovery 279
parent partition backup and recovery 279
Hyper-V child partition
configurations 274
365
Index
backup and recovery types 273
Hyper-V storage and backup options 274
I
incremental backup level
instant backup
overview 34
instant recovery
overview 36
instant snapshot policy example 57
interface
overview 39
issues
NMM client 340
items
searching for recovery 44, 234
L
label
creating for pool 53, 313
large SharePoint farms 175
LCR 195
Exchange backup 195
LCR backup support 195
level 1 to 9 backups
level for backup
considerations for 58, 314
Local Continuous Replication (LCR) 195
log files 337
NetWorker Server 349, 350
NMM client 347
PowerSnap client 347
Replication Manager 348
Solutions Enabler 348
troubleshooting with 347
NetWorker Deduplication Node, DNS resolution 75
NetWorker events
email notification 337
NetWorker Monitor window 337
Networker recovery options 89, 91
NetWorker Server
log files 349, 350
NetWorker server
connecting to 43
NMM Client
NMM client
error messages 338
issues 340
log files 347
nonauthoritative restore
ADAM 94
Cluster Writer 299
DFS 95, 96
FRS 97
NTDS 99
nonpersistent backup
overview 34
notification
NetWorker events 340
savegroup failed 340
notifications 337
types 337
NSR_ALT_PATH command 200
NSR_DATA_MOVER command 65
NSR_DM_PORT command 200
nsr_moss_save 186
nsr_sa_ini.exe 327
NTDS recovery
booting in directory services restore mode 102
consideration for 87
options 98
NUMBER_OF_PROCESSOR 175
O
M
marking items for recovery 44, 233
metadata recovery 174
Microsoft Cluster Service 296
Microsoft VSS software provider
overview 29
Microsoft Windows Server 2008 31, 62, 91, 93, 250, 253,
296, 308
Microsoft Windows Server 2008 support 91
mminfo, querying deduplication save sets using 80
monitoring a recovery 86, 102, 305
mount path for proxy client 200
MSCS 296
MSDE
turning writer on or off in registry 253
multiple client resources for backup 61, 256, 286, 315
366
N
overview
conventional recovery 36
Exchange proxy client 201
instant backup 34
instant recovery 36
interface 39
nonpersistent backup 34
persistent backup 34
provider 29
proxy client 34, 118
requestor 29
rollback recovery 36
rollover backup 34
serverless backup 34, 118, 201
software provider 29
VSS writers 30
overwrite prevention error
Exchange database recovery 342
Index
P
parent partition
pass phrase
recovering AES encrypted data 100
passive node 222, 223
password
changing 333
consideration for changing 333
restriction for active directory recovery 322
specifying for backup 333
password-protected
recovering backup data 86
persistent backup
overview 34
physical server
Hyper-V 274
policies for snapshots 55
pool
creating for snapshots 52, 312
label for 53, 313
ports
Exchange proxy client communications 200
service and connection port ranges 332
PowerSnap client
log files 347
prerequisites
deduplication 73
provider
overview 29
proxy client
configuring 67
mount path 200
overview 34, 118
port number for communicating 200
requirements for 67
SharePoint serverless backup 139
specifying hostname of 65
SQL serverless backup 118
SQL serverless backup in SharePoint 2003 156
R
recover.exe 86
recovery 179
ADAM options 93
browse time selection 45
Cluster options 91
Cluster Writer options 91
default method, choosing 90
DFS options 94
encrypted backup data 86
Exchange options 219
FRS options 97
Hyper-V 288
marking items 44, 233
NTDS options 98
password-protected backup data 86
performing for system data 87
searching for items 44, 234
selecting a version 45
SharePoint 2007 granular 173
SharePoint 2007 restore to an alternate location 183
snapshots that have not been rolled over 101
viewing progress for 86, 102, 305
volumes required for 46
recovery error
Exchange database 342
Recovery Storage Group (RSG)
Exchange 226
recovery, deduplication 80
remote access attribute
virtual cluster server specification 85, 257, 317
replica
error message 338
takes long time in CLARiiON environment 342
Replication Manager
error message 338
log files 348
requestor
overview 29
resource for backup
configuring 61, 315
rollback recovery
limitations of 327
overview 36
performing 327, 329
requirements for 327
rollover backup
device for 52, 312
overview 34
performing 326
S
sa.ini 327, 328
safe mode, booting in 102
save sets
applications 61
DPM 252
Exchange 198, 208, 211, 213, 214, 216
exclusion representation 40
Hyper-V 281, 282
SharePoint 2003 154
SharePoint 2007 137
SharePoint 2007 granular backup 186
specifying for backup 84, 256, 316
specifying for Hyper-V backup 286
SQL 115
SQL databases in SharePoint 2007 170
URL encoding 63
savegroup failed notification 340
Search Index
backup in SharePoint 2007 137
searching for a recovery item 44, 234
Security Configuration Wizard 340
selecting items for recovery 44, 233
server
connecting to NetWorker 43
serverless backup 32, 118, 156
367
Index
Exchange 201
overview 34
proxy client requirement 67
SharePoint 2007 backup 139
SQL 118
service port range
specifying 332
services 350
SharePoint 2003
application save sets 161
backup and recovery process 163
backup and recovery support 152
backups 157
Client resource 157
configuring Client resource 157
recovery 159
save sets 154
SQL 2005 database recovery 159
SQL Server 2000 recovery 159
SharePoint 2007
configuration database 129
configuration database backup 135
configuration database recovery 149
directed recovery through restore to an alternate
location 183
granular backup save sets 186
granular backup storage requirements 133
granular recovery prerecovery options 182
save sets for granular backup 176
Search Index 135
serverless backup 139
specifying save sets for backup 143, 145, 147
URL encoding 138
snapshot policies
configuring 55
examples of 57
snapshots
deleting 330
pool for 52, 312
recovering when not rolled over 101
rollback operation 327
rollover operation 326
software provider
overview 29
Solutions Enabler
log files 348
SQL
database recovery for SharePoint 2003 159
serverless backup 118
serverless backup in SharePoint 2003 156
URL encoding 117
URL encoding for SharePoint 2003 155
SQL proxy client 118
SQL proxy client in SharePoint 2003 backup 156
368
SQL Server
configure client resource 119
storage and backup options
Hyper-V 274
storage array
creating descriptor 328
descriptor file 327
Exchange rollback requirements 225
storage requirements
SharePoint 2007 granular backup 133
Symmetrix
provider overview 32
syntax
application save sets 61
DPM save sets 252
Exchange save sets 198
Hyper-V save sets 282
SharePoint2007 137
SQL 115
system
system component writers
supported 31
SYSTEM COMPONENTS
system recovery
performing 87
T
time of backup
troubleshooting
checking other documents 351
U
URL encoding
DPM save sets 253
Exchange save sets 199
save sets 63
SharePoint 2007 138
SharePoint 2007 save sets 138
SQL save sets 117
SQL save sets for SharePoint 2003 155
SQL save sets in SharePoint 2003 backup 155
V
version
virtual cluster server
entering physical hostname of 85, 257, 317
scheduling backups for 69, 298
Volume Shadow Copy Service (VSS)
writers 30
volumes required for recovery 46
VSS CLIENT... Invalid Writer Selection 339
VSS_E_VOLUME_NOT_SUPPORTED_BY_PROVIDER
341
VSS_E_WRITERERROR_RETRYABLE 338
Index
W
Windows Encrypting File System (EFS) 100
Windows Firewall
NetWorker blocked by 340
Windows Server 2003
Windows Server 2008 31, 62, 91, 93, 250, 253, 308
Windows Server 2008 Cluster Writer
Windows Server Failover Clustering (WSFC) 296
see SharePoint 2003 152
see SharePoint 2007 128
Windows system component writers
supported 31
369
Index
370

EMC NetWorker Module for Microsoft Applications

Transcription

Similar documents

Microsoft Windows Server Codename "Longhorn"

Issue 122 - Jan 2002 - Isle of Wight PC User Group

Ziua doctoranzilor

How To Use Rsync Backup Module

SVR302: Windows Server code name “Longhorn” Technical Overview

Request for Taxpayer ldentification Number and

Jacksonville Jaguars Select Cloud-Based Data Protection

Disk to Disk Backup for IBM Midrange Servers

Khalil Abd El

Blackadders uses FloSuite Legal