nmm_22_admin

EMC CorporationCorporate Headquarters:

Hopkinton, MA 01748-9103

1-508-435-1000www.EMC.com

EMC® NetWorker®

Module for Microsoft ApplicationsRelease 2.2

Administration GuideP/N 300-008-851

REV A02

EMC NetWorker Module for Microsoft Applications Release 2.2 Administration Guide 2

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.

Contents

Preface

Chapter 1 Overview NetWorker Module for Microsoft Applications................................................. 20

NetWorker Module features ........................................................................... 20Privileged user-level access............................................................................. 21Windows and Microsoft application updates .............................................. 21

Supported configurations...................................................................................... 22Supported NetWorker Server hosts............................................................... 22Windows operating system support ............................................................. 22Simple LAN-based configuration .................................................................. 23LAN-based configuration with a proxy client ............................................. 24LAN-free configuration ................................................................................... 25

The VSS snapshot creation process ...................................................................... 27Writers................................................................................................................ 28Providers............................................................................................................ 30Dynamic volume support ............................................................................... 30Providers and backup performance .............................................................. 31

NMM client snapshot backups ............................................................................. 32Nonpersistent rollover backup....................................................................... 32Instant backup with or without rollover....................................................... 32Serverless backup ............................................................................................. 32

NMM client recoveries........................................................................................... 34Conventional recovery..................................................................................... 34Instant recovery ................................................................................................ 34Rollback recovery ............................................................................................. 34Disaster and granular backup and recovery for applications.................... 35

The NMM client interface...................................................................................... 37Recover view ..................................................................................................... 37Snapshot Management view........................................................................... 39Monitor view..................................................................................................... 40

NMM client graphical user interface ................................................................... 41Opening the software and connecting to a NetWorker server .................. 41Selecting an item for recovery ........................................................................ 42Searching for an item ....................................................................................... 42Specifying a recovery browse time ................................................................ 43Selecting a backup version for recovery ....................................................... 43Viewing the volumes required for a recovery.............................................. 44


Contents

Chapter 2 Configuring a Scheduled Backup Backup configuration roadmap ............................................................................ 46

Privileged user-level access ............................................................................. 47Name resolution................................................................................................ 48Best practices and considerations for backing up application data........... 48

Task 1: Configure a backup pool for snapshot operations ................................ 50Configuring a device ........................................................................................ 50Configuring a label template........................................................................... 51Configuring a backup pool.............................................................................. 51Labeling the device ........................................................................................... 52

Task 2: Configure snapshot policies ..................................................................... 53Snapshot policies for SYSTEM COMPONENTS backup ............................ 53Creating a snapshot policy .............................................................................. 53Snapshot 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 ................................................................................................ 59Specifying save sets for application data....................................................... 59Application 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 ....... 67Cluster 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........... 71Avamar integration in NMM .......................................................................... 71Windows OS, application, and configuration support ............................... 72Unsupported for deduplication...................................................................... 72

Avamar and NetWorker server configuration .................................................... 73Configuring Avamar and the NetWorker server to receive deduplication data and process backups ...................................................... 73Setting DNS resolution for NetWorker Deduplication Node and NMM .......................................................................................................... 73

Deduplication backups........................................................................................... 74Backup levels and schedules for deduplication clients............................... 74Retention policies .............................................................................................. 74Backup configuration notes............................................................................. 75

Configuring a Client Resource for deduplication .............................................. 76Configuring a deduplication client ................................................................ 76

Recovering deduplicated data .............................................................................. 78Deleting deduplication save sets .................................................................... 78Querying deduplication save sets using mminfo ........................................ 78Additional NetWorker server deduplication operation information ....... 78

Chapter 4 System Backup and Recovery Windows system backup ....................................................................................... 80

Windows system support ................................................................................ 80

EMC NetWorker Module for Microsoft Applications Release 2.2 Administration Guide4

Contents

Specifying system save sets for recovery ...................................................... 80NMM System data commands ....................................................................... 81

Performing system backups.................................................................................. 82Task 5: Configure a System Client resource ................................................. 82

Performing a recovery............................................................................................ 84System recovery in Windows Server 2008 and Windows Server 2003..... 85System Recovery Options Summary ............................................................ 86Recovery options .............................................................................................. 87General recovery options ................................................................................ 87NetWorker recovery options .......................................................................... 87Cluster recovery options ................................................................................. 89ADAM recovery options ................................................................................. 91DFS recovery options....................................................................................... 92FRS recovery options ....................................................................................... 95NTDS recovery options ................................................................................... 96Security 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 .......................................................................................................... 108Disaster 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 ..................................................... 112

SQL Server versions supported by NMM................................................... 112Types of supported backup and recovery .................................................. 112Components used by NMM for SQL data backup and recovery............ 113Specifying save sets for SQL data ................................................................ 113Displaying valid SQL data save sets............................................................ 114URL encoding for SQL save sets .................................................................. 115SQL Server application information variable settings .............................. 116Serverless backup ........................................................................................... 116

Performing SQL Server backups ........................................................................ 117Full backup of an SQL Server ....................................................................... 117

Performing SQL Server recovery........................................................................ 119Recovering SQL Server 2005 and 2008 stand-alone databases from the SQLServerWriter save set.............................................................. 119Recovering SQL Server 2005 or 2008 express databases from the System Components save set ................................................................. 119Best practices for SQL Server backup and recovery.................................. 119

Performing SQL Server disaster recovery......................................................... 121SQL Server (standalone) disaster recovery on Windows Server 2008.... 121SQL Server (standalone) disaster recovery on Windows Server 2003.... 121SQL Server cluster disaster recovery on Windows Server 2008.............. 122SQL Server cluster disaster recovery on Windows Server 2003.............. 123


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 ............................ 127Protecting a SharePoint farm......................................................................... 128Storage requirements and performance impact in SharePoint backup .............................................................................................................. 129NMM installation requirements for a Microsoft Office SharePoint 2007 system ...................................................................................................... 129SharePoint 2007 data backed up by NMM.................................................. 131Keeping backups and recoveries in sync..................................................... 131Components used by NMM for SharePoint backup and recovery ......... 132Specifying SharePoint 2007 save sets for application data ....................... 133Displaying valid SharePoint 2007 data save sets ....................................... 133URL encoding for SharePoint save sets ....................................................... 134SharePoint 2007 application information variable settings ...................... 135

Performing SharePoint 2007 backups ................................................................ 136Configuring SharePoint 2007 backups......................................................... 136

Task 5: Configure a SharePoint 2007 Client resource ....................................... 138Performing a full backup for a distributed SharePoint farm.................... 143

Performing SharePoint 2007 recovery................................................................ 143Mandatory order for SharePoint recoveries................................................ 144Configuration database recovery requirements in full farm recovery ............................................................................................................ 144Selecting valid items for recovery ................................................................ 144Full recovery of a stand-alone SharePoint farm ......................................... 144Full recovery of a distributed SharePoint farm .......................................... 145Recovery of individual SharePoint content databases .............................. 146Recovery of SharePoint Search/Content Indexes ...................................... 146Rollback of SharePoint SQL databases ........................................................ 147

SharePoint 2003 backup and recovery ............................................................... 147Types of supported backup and recovery................................................... 147Components used by NMM for SQL backup and recovery ..................... 148Specifying save sets for SharePoint 2003 SQL data.................................... 148Displaying valid SQL data save sets for SharePoint 2003 backup........... 149URL encoding for SQL save sets in SharePoint 2003 backup................... 150SQL application information variable settings in SharePoint 2003 backup ..................................................................................................... 151

Performing SharePoint 2003 backups ................................................................ 152Task 5: Configuring a SharePoint 2003 Client resource ............................ 152

Performing a SharePoint 2003 recovery............................................................. 154Recovering SQL Server 2000 stand-alone databases for SharePoint 2003 ............................................................................................... 154Recovering SQL 2005 databases for SharePoint 2003 ................................ 154Best practices for SQL Server backup and recovery ofSharePoint 2003 databases ............................................................................. 156

SharePoint 2003 scenario of backup and recovery process ............................. 158 SharePoint 2007 disaster recovery ...................................................................... 165

Disaster recovery planning............................................................................ 165Performing a disaster recovery ..................................................................... 167


Contents

Chapter 7 Microsoft SharePoint 2007 Granular Backup and Recovery NMM support for granular backup and recovery of SharePoint items ....... 170

SharePoint objects available in NMM granular backup ........................... 170SharePoint limitations on document metadata recovery ......................... 170

Better performance for large SharePoint farms................................................ 171 Performing SharePoint 2007 granular backups................................................ 171 Configuring a SharePoint 2007 Client resource for granular backup ........... 172 SharePoint 2007 granular recovery .................................................................... 173 SharePoint recovery search ................................................................................. 175

Searching and marking tips .......................................................................... 177Searching for a SharePoint item ................................................................... 177

Performing a granular recovery of SharePoint items...................................... 178 SharePoint prerecovery options.......................................................................... 179

Selecting alternate destination options........................................................ 179 SharePoint recovery options ............................................................................... 181 Commands for SharePoint 2007 granular backup and recovery................... 182

nsr_moss_save.exe command....................................................................... 182nsr_moss_recover.exe command ................................................................. 186

Using administrator account other than that created during installation .............................................................................................................. 188

Chapter 8 Microsoft Exchange Server Backup and Recovery Exchange Server backup and recovery.............................................................. 190

Microsoft Exchange Server MAPI client and Collaboration Data Objects support...................................................................................... 190Exchange Server versions supported by NMM client .............................. 190NMM and Exchange Server support for L18N .......................................... 190Exchange Server backup and recovery types............................................. 190Updating and modifying Exchange credentials ........................................ 191Compatibility with previous Exchange Server backups .......................... 192Exchange Server data backed up by NMM ................................................ 193Components used by NMM for Exchange Server backup and recovery.................................................................................................... 193Specifying Exchange save sets for application data .................................. 194Exchange application information variable settings ................................. 196Serverless backup ........................................................................................... 197Snapshot process in an Exchange Server production and proxy client setup ........................................................................................... 198Best practices for Exchange backup and recovery..................................... 199

Performing Exchange Server backups............................................................... 202Configuring Exchange Server backups ....................................................... 202

Configuring an Exchange Client resource ........................................................ 204Configuring an Exchange Client resource .................................................. 204Configuring backup for Microsoft Exchange Server 2007 CCR .............. 206

Performing Exchange Server recovery .............................................................. 215Exchange recovery requirements and limitations ..................................... 215Quick recovery of a deleted mailbox ........................................................... 216Exchange recovery options ........................................................................... 217Roll-forward recovery.................................................................................... 218Point-in-time recovery ................................................................................... 219Mounting the database after recovery......................................................... 220Rollback restore of Exchange data by using NMM ................................... 220Exchange rollback operation requirements ................................................ 221

Recovery to storage groups ................................................................................ 222


Contents

Selecting a storage group restore destination............................................. 222Recovery to the original storage group ....................................................... 223Recovery to an Exchange RSG ...................................................................... 224Selecting and recovering items in the RSG database................................. 229Recovery to an alternate storage group....................................................... 231

Microsoft Exchange Server 2007 directed recovery.......................................... 234Performing a directed recovery .................................................................... 234

Microsoft Exchange Server 2007 disaster recovery .......................................... 237Microsoft Exchange Server 2007 (standalone) disaster recovery............. 237Microsoft Exchange Server 2007 disaster recovery for Exchange CCR to a 2 node cluster in a production environment.............................. 238Microsoft 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....................................... 245DPM backup and recovery types ................................................................. 246DPM data backed up by NMM..................................................................... 246Transportable backups of DPM replicas not supported ........................... 246Components used by NMM for DPM backup and recovery ................... 247Specifying DPM save sets for application data .......................................... 248DPM application information variable settings ......................................... 250

Performing DPM Server backups ....................................................................... 251Configuring 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 .................................................. 260Performing 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 ..................................................................... 268Operating system versions supported by NMM client ............................. 268Hyper-V parent partition backup and recovery types .............................. 269Hyper-V parent partitions, child partitions, and applications within child partitions.................................................................................... 269Hyper-V storage and backup options supported by NMM ..................... 270Backup roadmap for Hyper-V ...................................................................... 273Best practices for Hyper-V parent partition backup and recovery.......... 275Best practices for Hyper-V child partition applications backup and recovery ................................................................................................... 276Components used by NMM for Hyper-V backup and recovery ............. 277Specifying Hyper-V save sets for application data .................................... 277Hyper-V application information variable settings................................... 279

Performing Hyper-V backups ............................................................................. 280Preparing a virtual machine for backup...................................................... 280


Contents

Hyper-V configuration requirements for backing up a virtual machine that contains multiple volumes .................................................... 280NMM support for proxy backup of Hyper-V ............................................ 280Configuring Hyper-V backups..................................................................... 281

Configuring a Hyper-V Client resource ............................................................ 282 Performing Hyper-V recovery ............................................................................ 284

Off-line and on-line recovery........................................................................ 284Restrictions for backup and recover of Hyper-V virtual machines in a Windows 2008 Failover Cluster .......................................... 284Hyper-V recovery to the original machine and location .......................... 285Hyper-V recovery to a different machine or location ............................... 285

Hyper-V parent partition disaster recovery ..................................................... 288

Chapter 12 Microsoft Windows Cluster Backup and Recovery Windows Server Cluster backup and recovery................................................ 292

Cluster support in NMM client .................................................................... 292Compatibility with previous Windows Cluster backups......................... 292Components used by NMM for Windows Cluster backup and recovery.................................................................................................... 292Windows Cluster application information variable settings ................... 293

Performing Windows Server Cluster backups ................................................. 294Cluster failover and backups ........................................................................ 294

Windows Server Cluster recovery...................................................................... 295 Windows Server 2003 clustered client disaster recovery ............................... 295

Disaster recovery of a cluster client ............................................................. 295Disaster recovery from a legacy VSS backup ............................................. 298Disaster recovery of a clustered client from a legacy non-VSS backup .............................................................................................................. 300Monitoring a recovery ................................................................................... 301

Chapter 13 Microsoft Active Directory Backup and Recovery Active Directory backup and recovery.............................................................. 304

Types of supported backup and recovery .................................................. 304 Performing an Active Directory or ADAM backup ........................................ 305

Planning backups for disaster recovery ...................................................... 305 Performing a granular Active Directory or ADAM backup........................... 306

Considerations for granular backup............................................................ 306Configuring Active Directory or ADAM granular backups .................... 307Task 1: Set up local Windows SYSTEM account as an ADAM administrator................................................................................................... 307Task 2: Configure a pool for backup operations........................................ 308Task 3: Configure a backup schedule .......................................................... 310Task 4: Configure a backup group............................................................... 310Task 5: Configure a Client resource ............................................................. 311Task 6: Configure privileges ......................................................................... 313

Recovering all Active Directory data................................................................. 314 Performing a granular recovery of Active Directory objects and attributes ................................................................................................................. 316

Recovery Restrictions for Active Directory ................................................ 317 Active Directory disaster recovery..................................................................... 320


Contents

Chapter 14 Snapshot Management Roll over a snapshot.............................................................................................. 322 Roll back a snapshot ............................................................................................. 323

Rollback recovery requirements ................................................................... 323Storage array descriptor file .......................................................................... 323System rollback may fail with CLARiiON if the configuration information is not current.............................................................................. 325Performing 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 .................................................................................. 333Finding errors .................................................................................................. 333

NMM client error messages................................................................................. 334E_VOLUME_NOT_SUPPORTED ................................................................ 334E_VETO_PROVIDER ..................................................................................... 334VSS_E_WRITERERROR_RETRYABLE error code 0x800423f3................ 334VSS CLIENT... Invalid Writer Selection... for APPLICATIONS .............. 335access to the path is denied............................................................................ 335Object reference not set to an instance of an object.................................... 335

NMM client issues................................................................................................. 336Savegroup failed in scheduled backup........................................................ 336In progress backup fails ................................................................................. 336NetWorker is blocked by Windows Firewall.............................................. 336Need to run utility to obtain Exchange Server information ..................... 336Requesting a snapshot for a volume on hardware that is not transportable ............................................................................................. 337PowerSnap service must be restarted after the NetWorker serviceis restarted........................................................................................................ 337PowerSnap validation for recovery fails ..................................................... 338Snapshot backup may fail with writer errors ............................................. 338Microsoft I/O write error when taking a replica in a CLARiiON array .................................................................................................................. 338Creating a replica takes a long time in CLARiiON environment ............ 338Cannot recover Exchange database due to overwrite preventionerror................................................................................................................... 338Incremental backups within NetWorker Module for ExchangeClient may be promoted to full backups ..................................................... 339Incremental backups within NetWorker Module for SQL ServerClient may be promoted to full backups ..................................................... 339Diagnosing VSS Writer Issues....................................................................... 340Modifying the Exchange 2007 COM+ component’s user account or password ..................................................................................................... 340Save sets may not appear in the recover window...................................... 341Recovery of AES-encrypted files fails without passphrase ...................... 341Unable to restore files encrypted with both Microsoft Windows Encrypting File System and AES encryption.............................................. 341Anti-virus programs block recovery ............................................................ 342


Contents

Checking log files.................................................................................................. 343NMM client...................................................................................................... 343PowerSnap client ............................................................................................ 343Replication Manager ...................................................................................... 344Solutions Enabler............................................................................................ 344Active Directory.............................................................................................. 345NetWorker server ........................................................................................... 345

Manually stopping and starting services .......................................................... 346 Other troubleshooting resources ........................................................................ 347

Glossary

Index


Contents


Title Page

Figures

1 Simple LAN-based configuration ....................................................................................... 262 LAN-based configuration with a proxy client .................................................................. 273 LAN-free configuration........................................................................................................ 284 VSS backup process .............................................................................................................. 305 NMM client main window .................................................................................................. 396 Exclusion representation of recover items ........................................................................ 407 Snapshot Management view ............................................................................................... 418 Monitor view ......................................................................................................................... 429 Selected and partially selected items.................................................................................. 4410 Local client on the taskbar next to Client......................................................................... 10611 Configuration Options dialog box.................................................................................... 10612 Select Viewable Clients dialog box................................................................................... 10713 NMM client dropdown list ................................................................................................ 10714 SharePoint 2007 stand-alone farm configuration ........................................................... 13415 SharePoint 2007 distributed farm configuration ............................................................ 13516 Restore Remote Dependencies dialog box ...................................................................... 15017 Newly created databases created in SharePoint............................................................. 16418 NMM backups available for recovery.............................................................................. 16519 SQL databases selected for recovery in NMM................................................................ 16520 Databases recovered to SQL Server.................................................................................. 16621 SharePoint Server Central Administration window...................................................... 16622 SharePoint Portal Server Central Administration, Manage Portal Sites page ........... 16723 Portal creation and restore options................................................................................... 16724 Portal restore and database names ................................................................................... 16825 Confirmation page for request to create a portal site .................................................... 16826 Status page for portal creation process ............................................................................ 16927 Operation Successful page displayed when portal creation is complete.................... 16928 Recovered databases and SharePoint portal visible and available in SharePoint ..... 16929 NMM displaying SharePoint 2007 browsing for recovery items................................. 17930 NMM displaying Search tab for SharePoint 2007 recovery.......................................... 18031 Sample output from nsr_moss_save ................................................................................ 18732 Snapshot process with Exchange production server and proxy client ....................... 20333 Select Storage Group Restore Destination dialog box ................................................... 22634 Available and unavailable storage groups. ..................................................................... 23035 Restore To command on context menu ........................................................................... 23036 RSG Configuration error message .................................................................................... 23137 RSG Configuration correct message................................................................................. 23138 Public folders cannot be recovered to an RSG................................................................ 23239 Selected and partially selected items................................................................................ 23440 Recovered items folders in user mailbox......................................................................... 235


41 Local client on the taskbar next to Client ......................................................................... 23942 Configuration Options dialog box .................................................................................... 23943 Select Viewable Clients dialog box ................................................................................... 23944 NMM client dropdown list ................................................................................................ 24045 NMM protection of DPM ................................................................................................... 24946 Directed granular recovery to a DPM-protected client.................................................. 26047 DPM recovery modes.......................................................................................................... 26148 Recover submenu for DPM Recovery .............................................................................. 26149 DPM granular recovery available to file level................................................................. 26250 DPM granular recovery options........................................................................................ 26251 Directed recovery of FileServer1 ....................................................................................... 26452 DPM database and replicas available as selectable items.............................................. 26753 DPM recovery modes.......................................................................................................... 26754 Recover submenu for DPM recovery ............................................................................... 26855 DPM disaster recovery options ......................................................................................... 26856 Windows Server 2008 host with Hyper-V virtual machine child partitions .............. 27457 Hyper-V storage options .................................................................................................... 27558 NMM backup of Hyper-V components ........................................................................... 28159 Disaster recovery order for a Hyper-V server................................................................. 29260 Troubleshooting process..................................................................................................... 336


Title Page

Tables

1 Currently supported writers ............................................................................................... 312 Support for disaster and granular backup and recovery ................................................ 383 Backup tasks .......................................................................................................................... 484 Access privileges needed for NMM ................................................................................... 495 Best practices and considerations for application backups ............................................ 506 Conventional backup settings ............................................................................................. 577 Instant snapshot policy ........................................................................................................ 578 Instant snapshot with rollover policy ................................................................................ 579 Considerations for NMM Client backup schedules......................................................... 5810 Save set syntax....................................................................................................................... 6211 Special characters and their URL-encoded values ........................................................... 6412 Application information variable settings......................................................................... 6513 Configuration requirements for a proxy client................................................................. 6714 Additional steps and considerations for backing up a clustered Client resource ....... 6915 NMM System data commands............................................................................................ 8316 Access privileges needed for NMM ................................................................................... 8617 SQL save set syntax............................................................................................................. 11518 Special characters and their URL-encoded values ......................................................... 11719 SQL variable settings for Application Information attribute ....................................... 11820 SQL Server backup and recovery best practices and considerations .......................... 12221 Comparison of VSS, optimized, and granular backups ................................................ 13122 SharePoint save set syntax................................................................................................. 13723 Special characters and their URL-encoded values ......................................................... 13924 SharePoint application information variable settings ................................................... 13925 Save sets for backup............................................................................................................ 14026 SQL save set syntax for SharePoint 2003 ......................................................................... 15327 Special characters and their URL-encoded values ......................................................... 15528 SQL variable settings for application information attribute......................................... 15629 SQL Server backup and recovery best practices and considerations .......................... 16130 SharePoint Server backup list for disaster recovery ...................................................... 17031 SharePoint recovery search fields ..................................................................................... 18132 SharePoint granular recovery save set syntax for SharePoint 2007 objects................ 18633 Application information variable settings....................................................................... 18934 nsr_moss_recover command options............................................................................... 19035 Exchange save set syntax ................................................................................................... 19836 Special characters and their URL-encoded values ......................................................... 19937 Exchange application information variable settings...................................................... 20038 NMM Data Protection Manager commands................................................................... 25139 DPM Server 2007 save set syntax...................................................................................... 25240 DPM application information variable settings ............................................................. 254


41 Hyper-V child partition configurations ........................................................................... 27542 Backup tasks for Hyper-V .................................................................................................. 27743 NMM Hyper-V commands ................................................................................................ 28144 Hyper-V save set syntax ..................................................................................................... 28245 Hyper-V application information variable settings........................................................ 28346 Windows Cluster application information variable settings ........................................ 29747 Additional steps and considerations for backing up a clustered Client resource ..... 29848 Considerations for NMM client backup schedules ........................................................ 31449 System-only attributes that are not backed up................................................................ 32150 Retained attributes after object is deleted ........................................................................ 32251 Notifications and errors...................................................................................................... 33752 Services used in NMM client ............................................................................................. 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:

◆ Storage subsystems, such as EMC CLARiiON or Symmetrix, if used

◆ EMC NetWorker software

◆ EMC NetWorker Snapshot management

◆ Microsoft Volume Shadow Copy Service (VSS) technology

Relateddocumentation

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

Conventions EMC uses the following conventions:

Note: A note presents information that is important, but not hazard-related.


Preface

CAUTION

A caution note contains information that is essential to avoid data loss.

Typographical conventionsEMC uses the following type style conventions in this document:

Where to get help 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.

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

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


http://powerlink.emc.com

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]


http://softwaresupport.EMC.com

Preface


1

This chapter includes the following topics:

◆ NetWorker Module for Microsoft Applications ....................................................... 22◆ Supported configurations ............................................................................................ 24◆ The VSS snapshot creation process............................................................................. 29◆ NMM client snapshot backups ................................................................................... 34◆ NMM client recoveries ................................................................................................. 36◆ The NMM client interface ............................................................................................ 39◆ NMM client graphical user interface.......................................................................... 43

Overview

Overview 21

Overview

NetWorker Module for Microsoft ApplicationsThe 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 featuresThis 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


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 configurationsThis section describes the basic configurations in which the NMM client is supported.

Supported NetWorker Server hostsThe 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:

◆ 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.

Supported configurations 25

Overview

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.



LAN

NetWorkerStorage

Node 7.3.x

Primary storage(For example,

CLARiiON or Symmetrix)

Traditionalstorage

NetWorkerServer 7.3.3 or later

Application serverNetWorker Module forMicrosoft Applications

SAN

GEN-000747

S1 S2

S3S4

S5


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.

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 where the NMM client software is installed.



LAN

NetWorkerStorage

Node 7.3.xPrimary storage

(For example,CLARiiON or Symmetrix)

Traditionalstorage

S1 S2

S3S4

S5

NetWorker Server7.3.3 or later


Proxy clientNetWorker Module forMicrosoft Applications

SAN

GEN-000748

Supported configurations 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.

Figure 3 LAN-free configuration

LAN

Primary storage(For example,

CLARiiON or Symmetrix)

Traditionalstorage

S1 S2

S3S4

S5



Proxy clientNetWorker StorageNode 7.3.x

SAN

GEN-000750


Overview

The VSS snapshot creation processMicrosoft Volume Shadow Copy Service (VSS) acts as a coordinator among the following three components that create, modify, back up, and recover data:

◆ RequestorsRequestors include the NetWorker backup application.

◆ WritersIncluding 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.

◆ ProvidersProviders 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.

The VSS snapshot creation process 29

Overview

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.

Writer Writer

Writer Writer

Writer

Requestor

4. Requestor asks VSS to create snapshot.

1. NetWorker software (the requestor) asks VSS to enumerate writers and gether their metadata.

2. Writers provide an XML description of backup components and define the recover method.

5. VSS tells the writers to freeze activity.

7. VSS tells the writers to resume activity.

3. VSS asks which providers can support a snapshot for each of the required volumes.

6. VSS tells the providers to create the snapshot of the current state on disk.

Disk 1 Disk 2 Disk 3

Microsoft VSSSoftware Provider

EMC VSS Provider(CLARiiON or Symmetrix)

Disk 3GEN-000188


Overview

Table 1 on page 31 lists the writers that are supported for the NMM client.

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.

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 2003Microsoft Exchange Server 2007 Microsoft System Center Data Protection Manager 2007Microsoft Distributed File Replication (Windows Server 2003)Microsoft Active Directory Lightweight Directory Services (AD LDS)Microsoft SharePoint Portal Server 2003Microsoft Office SharePoint Server 2007Windows SharePoint Services 2.0Windows SharePoint Services 3.0Microsoft SQL Server 2008 Microsoft SQL Server 2005Windows Server 2008 (Full Installation) with Hyper-VWindows 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 AuthorityCluster Database (Windows Server 2003)Cluster Writer (Windows Server 2008)DHCP (Dynamic Host Configuration Protocol) JetEvent logFRS (File Replication Server)FSRM (File Storage Resource Manager)Microsoft Distributed File Replication (Windows Server 2003)IIS Configuration Writer (Windows Server 2008)IIS MetabaseMSSearch Service Writer (Windows Server 2008)NPS VSS Writer (Network Policy and Access Service, Windows Server 2008)NTDS (Active Directory Schema)RegistryRSM (Removable Storage Manager)Remote Storage ManagerShadow Copy Optimization Writer (Windows Server 2008)SQL Embedded (SQL 2005 Express, or SQL 2008 Express)Terminal Services Gateway Writer (Windows Server 2008)Terminal Services LicensingWINS (Windows Internet Name Service)WMI (Windows Management Instrumentation)


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 supportIn 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.


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.


Overview

NMM client snapshot backupsBackups 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 backupA 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


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 recoveriesThere 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 recoveryA 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 recoveryInstant 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.


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 Support for disaster and granular backup and recovery

Active Directory

SQL Server

Exchange Server

Microsoft Office

SharePoint Server 2007

Data Protection Manager Hyper-V

File system

Disaster (full backup)

X X X X X X X

Granular backup

X X X X X

Disaster (full recovery)

X X X X X X X

Granular recovery

X X X X X


Overview

The NMM client interfaceThe 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.

Figure 5 NMM client main window

Navigation tree

View-specific toolbarAvailable views

Client selection (clustered client only)

Main toolbar

Summary of selected items

The NMM client interface 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.

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.

Items in red are excluded from recovery


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.

The NMM client interface 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.


Overview

NMM client graphical user interfaceThis section describes how to perform the basic tasks in the NMM client interface.

Opening the software and connecting to a NetWorker serverYou 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.

NMM client graphical user interface 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.

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.

Check mark indicates that the item is selected.Solid box indicates that some items, not all, below this node are selected


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.

NMM client graphical user interface 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 for recovery” on page 44 provides more information about 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.


2


◆ Backup configuration roadmap .................................................................................. 48◆ Task 1: Configure a backup pool for snapshot operations ...................................... 52◆ Task 2: Configure snapshot policies ........................................................................... 55◆ Task 3: Configure a backup schedule ......................................................................... 58◆ Task 4: Configure a backup group .............................................................................. 59◆ Task 5: Configure a Client resource ............................................................................ 61◆ Task 6: Configure privileges ........................................................................................ 66◆ Task 7: Configure a proxy client.................................................................................. 67◆ Backing up a clustered NMM Client .......................................................................... 69◆ Moving an NMM Client to another NetWorker server ........................................... 70

Configuring aScheduled Backup

Configuring a Scheduled Backup 47

Configuring a Scheduled Backup

Backup configuration roadmapThis 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.

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:\”

• 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

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.

• Complete tasks 1 through 4:“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

• 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 SharePoint 2007 backups” on page 141“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



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.

For all other applications and systems supported by NMM


An NMM client that is part of a cluster • Review the information in “Backing up a clustered NMM Client” on page 69


An Active Directory server Complete the tasks in Chapter 13, “Microsoft Active Directory Backup and Recovery.”

Table 3 Backup tasks (page 2 of 2)

Type of backup Required action

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 ServerMicrosoft Data Protection ManagerMicrosoft Office SharePoint ServerMicrosoft SQL ServerMicrosoft Exchange Server

Windows Adminstrator

Network domain Windows Domain Administrator

Storage node servers Windows Administrator

Backup configuration roadmap 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 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 dataAssign 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.



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

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.

Microsoft SQL Server 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.

Table 5 Best practices and considerations for application backups (page 2 of 2)


Backup configuration roadmap 51


Task 1: Configure a backup pool for snapshot operationsSnapshots 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 deviceConfigure 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.



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:


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:

1. In the Administration window of the NetWorker Management Console, click 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 2: Configure snapshot policiesA 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:

1. In the Administration window of the NetWorker Management Console, click 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.

Task 2: Configure snapshot policies 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!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.

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.



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 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 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 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 Instant snapshot policy

Name Instant backup

Comment Persistent snapshots with no rollovers


Retain snapshots 28

Snapshot expiration policy Week

Backup snapshots None

Table 8 Instant snapshot with rollover policy

Name Instant and conventional backup

Comment Persistent snapshots with rollover to tape


Retain snapshots 28

Snapshot expiration policy Week

Backup snapshots Every 1 Every 3

Task 2: Configure snapshot policies 57


Task 3: Configure a backup scheduleBackup 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.

Note: The information in Table 9 on page 58 applies only to rollover backup operations. Snapshots are always taken at full level.

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\.



Task 4: Configure a backup groupBackup 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:


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.

Task 4: Configure a 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.



Task 5: Configure a Client resourceA 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 applicationsThe 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 Exchange Server backups” on page 206

◆ “Performing DPM Server backups” on page 255

◆ “Performing SharePoint 2003 backups” on page 157

◆ “Performing SharePoint 2007 backups” on page 141

◆ “Performing SharePoint 2007 granular backups” on page 175

◆ “Performing SQL Server backups” on page 119

◆ “Performing Hyper-V backups” on page 284

◆ “Performing Windows Server Cluster backups” on page 298

Specifying save sets for application dataTable 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.

Task 5: Configure a Client resource 61


Table 10 Save set syntax

Type of backup data Save set syntax

DFS data (Windows Server 2003 only) APPLICATIONS:\DFS Replication service writerAlso 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:\.

ADAM data APPLICATIONS:\ADAM (instance_name) WriterFor example, the ADAM instance named AddressBook would be typed as:APPLICATIONS:\ADAM (AddressBook) WriterChapter 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.

Microsoft Exchange Server “Displaying valid Exchange data save sets” on page 198 provides more information about Exchange save sets and syntax.

Microsoft SQL Server “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.

Microsoft Office SharePoint 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 setsTo 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""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 setsWhen 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



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 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.

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

Table 12 Application information variable settings

Attribute name Description Values

NSR_SNAP_TYPE=value Specifies the snapshot service provider name.

vssThis 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 hostnameIf 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.



Task 6: Configure privilegesThe 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:


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.



Task 7: Configure a proxy clientA 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


Additional requirements for an Exchange Server

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.

Table 13 Configuration requirements for a proxy client (page 2 of 2)

Requirement Description



Backing up a clustered NMM ClientTo 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.

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.

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.

Backing up a clustered NMM Client 69


Moving an NMM Client to another NetWorker serverDo 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.


3


◆ Data deduplication........................................................................................................ 72◆ Avamar and NetWorker server configuration .......................................................... 75◆ Deduplication backups................................................................................................. 76◆ Configuring a Client Resource for deduplication .................................................... 78◆ Recovering deduplicated data..................................................................................... 80

Deduplication Backupand Recovery

Deduplication Backup and Recovery 71

Deduplication Backup and Recovery

Data deduplicationThis 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:

◆ 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

◆ Windows SharePoint Services 3.0

◆ Windows SharePoint Services 2.0

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

◆ Windows configurations: EFS File Systems



Avamar and NetWorker server configurationBefore 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 backupsAll 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.



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 deduplicationWhen 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 clientAfter 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.



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 mminfoYou 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.


4


◆ 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

System Backup andRecovery

System Backup and Recovery 81

System Backup and Recovery

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 supportNMM 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 recoveryYou 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:

◆ 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)

◆ Microsoft Distributed File Replication (Windows Server 2003)

◆ 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 backupsThe 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

◆ “Task 2: Configure snapshot policies” on page 55

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.

◆ “Task 3: Configure a backup schedule” on page 58

◆ “Task 4: Configure a backup group” on page 59

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:


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.



• 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!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.

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 recoveryMake 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.

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.

◆ 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:\.

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 ServerMicrosoft Data Protection ManagerMicrosoft Office SharePoint ServerMicrosoft SQL ServerMicrosoft Exchange Server

Windows Administrator

Network domain Windows Domain Administrator

Storage node servers Windows Administrator



◆ 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 dataThe 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.

Performing a recovery 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.



“Performing a recovery” on page 86 provides more information.

Recovery optionsGeneral 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 optionsThis 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 optionsThe 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.



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.



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 restoreUse 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.



6. Select the Authoritative Restore options.


8. In the navigation tree, expand the SYSTEM COMPONENTS folder and select 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 upgradeIf 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.



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 restorePerform 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.



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 optionsYou 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:

◆ 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.

Nonauthoritativerestore (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:

1. From the Options menu, select System Recover Session Options and then click 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.




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.

“Windows system backup” on page 82 provides more information about recovering system data.

To specify which type of DFS recovery to perform in Windows Server 2008:

1. From the Options menu, select System Recover Session Options and then click 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.




FRS recovery options

You can specify whether to perform an authoritative or nonauthoritative recovery of 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 recoveryAn 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 specifying save sets when configuring a backup.

To specify whether to perform an authoritative or nonauthoritative recovery:

1. From the Options menu, select System Recover Session Options and then click 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.





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.


NTDS recovery options

You can specify whether to perform an authoritative or nonauthoritative recovery of 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 specifying save sets when configuring a backup.



To specify whether to perform an authoritative or nonauthoritative recovery:

1. From the Options menu, select System Recover Session Options and then click 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.



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.




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:

1. From the Options menu, select System Recover Session Options and then click 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.



Recovering file system snapshots that have not been rolled overThe 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 an item for recovery” on page 44 provides more information about 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 modeBefore 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:

◆ “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 dataPerform 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.



Performing a directed recoveryIn 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 NMMTo 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.

Performing a directed recovery 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.

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.

Refresh button



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.

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.

Client list

Performing a directed recovery 107


Recovering the Windows system configuration to an earlier stateHosts 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.



Disaster recoveryNMM 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.



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

5


◆ Microsoft SQL Server backup and recovery............................................................. 114◆ Performing SQL Server backups................................................................................ 119◆ Performing SQL Server recovery .............................................................................. 121◆ Performing SQL Server disaster recovery ............................................................... 123

Microsoft SQL ServerBackup and Recovery

Microsoft SQL Server Backup and Recovery 113

Microsoft SQL Server Backup and Recovery

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 NMMNMM 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.

The EMC Information Protection Software Compatibility Guide provides the most 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:

◆ 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 stand-alone databases



◆ 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 stand-alone databases

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%5Cinstance\masterFor example, to back up the database named trans07 on the SQL Server named instance MT11\BU, type the following:APPLICATIONS:\SqlServerWriter\MT11%5CBU\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.

Microsoft SQL Server backup and recovery 115


Displaying valid SQL data save sets

To display a list of the SQL save sets that are available for backup:


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:



3. Press Enter.

The save sets that are available for backup are listed in a format similar to the following:

"APPLICATIONS:\SqlServerWriter""APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON""APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\master""APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\model""APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\msdb""APPLICATIONS:\SqlServerWriter\CORPDBSERV1%5CYUKON\NADATA"


All SQL Server 2005 named instances and databases, including express and stand-alone databases


Only the SQL 2005 host instance and database name combination

APPLICATIONS:\SqlServerWriter\host%5Cinstance\masterFor example, to back up the database named trans07 on the SQL Server named instance MT11\BU, type the following:APPLICATIONS:\SqlServerWriter\MT11%5CBU\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.

Table 17 SQL save set syntax (page 2 of 2)




URL encoding for SQL 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:



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


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.



\ %5C ? %3F

/ %2F ] %5D

" %22 [ %5B

% %25 } %7D

# %23 { %7B

& %26 ^ %5E

< %3C ‘ %60

> %3E | %7C

Microsoft SQL Server backup and recovery 117


SQL Server application information variable settings

Table 19 on page 118 lists the variables that can be specified in the Application Information attribute of the Client resource.

Serverless backupA 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.

A serverless backup can be rolled over to a conventional backup, which can remain on the mounted storage disk volume as a persistent snapshot.


Table 19 SQL variable settings for Application Information attribute



vssThis value is required.

NSR_DATA_MOVER=value application information attribute

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 hostnameIf 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 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.



Performing SQL Server backupsNMM 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 ServerThe 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 5: Configure a SQL Server Client resourceThe next task is specific to configuring a SQL Server Client Resource:








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:

APPLICATIONS:\SqlServerWriter

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.



Performing SQL Server recoveryMake 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 SQL Server backup and recovery best practices and considerations



For application servers such as SQL servers, consider backing 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 dataAssign the appropriate snapshot policy and Client resource to each backup group.




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 the EMC VSS Provider.

NetWorker modules and 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 2008Prior 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 2003Prior to disaster recovery, use NMM to perform regular backups of volumes, system components, and application save sets.



C:\

SYSTEM COMPONENTS:\

Performing SQL Server disaster recovery 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 2008Prior to disaster recovery, use NMM to perform regular backups of volumes, system components, and application save sets.

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:\









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 2003Prior to disaster recovery, use NMM to perform regular backups of volumes, system components, and application save sets.

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.

Performing SQL Server disaster recovery 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.


6


◆ SharePoint Server versions supported by NMM client ......................................... 128◆ Types of SharePoint Server 2007 backup and recovery ......................................... 129◆ SharePoint Server 2007 backup and recovery ......................................................... 133◆ About optimized backup ........................................................................................... 140◆ Performing SharePoint 2007 backups ...................................................................... 141◆ Task 5: Configure a SharePoint 2007 Client resource ............................................. 143◆ Performing SharePoint 2007 recovery...................................................................... 148◆ SharePoint 2003 backup and recovery ..................................................................... 152◆ Performing SharePoint 2003 backups ...................................................................... 157◆ Performing a SharePoint 2003 recovery................................................................... 159◆ SharePoint 2003 scenario of backup and recovery process ................................... 163◆ SharePoint 2007 disaster recovery ............................................................................ 170

Microsoft SharePointServer Backup and

Recovery

Microsoft SharePoint Server Backup and Recovery 127

Microsoft SharePoint Server Backup and Recovery

SharePoint Server versions supported by NMM clientThe EMC Information Protection Software Compatibility Guide provides the most up-to-date lists of hardware, operating system, service pack, and application versions supported by the 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 3.0Backup 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 2.0Backup 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.



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– ListChapter 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

Types of SharePoint Server 2007 backup and recovery 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.



◆ 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 backupOptimized 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

SharePoint disaster recovery Deduplication Global source based deduplication

X X X

Target based deduplication X X X

Topologies Traditional LAN X X X

Off-host via proxy X

LAN-free (SAN) X Proxy

Backup support Farm X

Databases X

Backup types Full X X X

Incremental (log) X

Differential

Recovery Farm X

CFG Database X

Search database/index X

Content databases X

Types of SharePoint Server 2007 backup and recovery 131


SharePoint granular recovery Recovery Web applications X X

Site collection X X

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.

Table 21 Comparison of VSS, optimized, and granular backups (page 2 of 2)

SharePoint data protection Feature Support provided VSS backupOptimized backup

Granular backup



SharePoint Server 2007 backup and recoveryThis 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 processIn 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.

SharePoint Server 2007 backup and recovery 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.

Figure 14 SharePoint 2007 stand-alone farm configuration

ConfigurationDatabase

ContentDatabase

NetWorker Server7.3.3 or later:

Administers protectionfor NetWorker clients

such as NMM

Web Front End

Index Server

NetWorker Module forMicrosoft Applications

GEN-000751

Microsoft SQL Server Query Server



Figure 15 SharePoint 2007 distributed farm configuration

SharePoint 2007 data backed up by NMMNMM 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 syncSet 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.

Configuration Database(1 per farm)

Content Database 1(1 or more per farm)

Content Database 2

GEN-000752

Query ServerIndex ServerIf there is more than QueryServer, the Query Serverscannot be included on theIndex ServerNetWorker Module forMicrosoft Applications

Microsoft SQL ServerContains all SharePoint

databasesNetWorker Module forMicrosoft Applications

NetWorker Server7.3.3 or later:Administers

protection forNetWorker clients

such as NMM

Web Front EndOn an application server

NetWorker Module forMicrosoft Applications



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:

◆ 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.

Displaying valid SharePoint 2007 data save setsTo 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:


Note: If the application server is on a virtual host, run this command from the physical node that currently hosts the application server.

Table 22 SharePoint save set syntax


SharePoint databases on SQL Server


Full SharePoint backupThe SharePoint Search Index and SharePoint Content Index are included when this save set is specified.

APPLICATIONS:\Microsoft Office SharePoint Services

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.

SharePoint content database A specific content database.For example, on the host Mars1, a database named generic_database_1

APPLICATIONS:\Microsoft Office SharePoint Services\Mars1\generic_database_1

SharePoint Search Index and SharePoint Content Index

SYSTEM COMPONENTS:\Microsoft Office SharePoint Searchor SYSTEM COMPONENTS:\Microsoft Office Search

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.



3. Press Enter.




URL encoding for SharePoint 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 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:

APPLICATIONS:\Microsoft Office SharePoint Services\Mars1%5generic_database_1



SharePoint 2007 application information variable settings




\ %5C ? %3F

/ %2F ] %5D

" %22 [ %5B

% %25 } %7D

# %23 { %7B

& %26 ^ %5E

< %3C ‘ %60

> %3E | %7C

Table 24 SharePoint application information variable settings




NSR_DATA_MOVER=value application information attribute

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 hostnameIf 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.



About optimized backupOptimized 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 recoveryFor optimized backup, the backup can be scheduled using the NMC GUI, or started manually from command line.

By using the NMC GUITo 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.

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.

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.



Command line backup or manual backupThe 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 backupsNMM 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 backupsThe 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.

Performing SharePoint 2007 backups 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:

◆ “Task 5: Configure a SharePoint 2007 Client resource” on page 143



Task 5: Configure a SharePoint 2007 Client resourceFor 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 farmIn 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:









9. In the Save Set attribute, specify the components to be backed up. Place multiple 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.

Task 5: Configure a SharePoint 2007 Client resource 143


10. 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 SharePoint save sets” on page 138.

Note: If a save set is copied from the nsrsnap_vss_save - ? output, it will already include the URL encoding.


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.







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 not be able to display backups.

19. Click OK.

Configuring a SharePoint Client resource to perform a full backup of a distributed farmIn 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.









8. From the Retention Policy attribute, select a retention policy from the list. The 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.


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: If a save set is copied from the nsrsnap_vss_save - ? output, it will already include the URL encoding.




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 farmTo configure a SharePoint Client resource to back up a component on a machine in a distributed farm:












• 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 .

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.













NSR_SNAP_TYPE=vss


These settings are described in “SharePoint application information variable settings” on page 139.






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 recoveryAfter the backup is complete, NMM truncates the files. When the files are backed up, administrators can perform:

◆ “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.

Performing SharePoint 2007 recovery 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.



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:

APPLICATIONS:\Microsoft Office SharePoint 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.


The Restore Remote Dependencies dialog box appears, alerting you to go to the 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.


Recovery of SharePoint Search/Content IndexesTo perform a full recovery of Microsoft Office SharePoint Server 2007:

1. Open NetWorker Module for Microsoft Applications.

2. In the navigation tree, expand the SYSTEM COMPONENTS folder and select SPSearch VSS Writer and OSearch VSS Writer.


The Restore Remote Dependencies dialog box appears, alerting you to go to the specified hosts and start the recovery process.

Performing SharePoint 2007 recovery 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.


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 recoveryMicrosoft 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.

Types of supported backup and recovery 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.

◆ 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)


All SQL Server 2005 named instances and 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

SharePoint 2003 backup and recovery 153


Displaying valid SQL data save sets for SharePoint 2003 backupTo 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-SQL2000APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServerAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServerTempDBAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5CSPS01_Config_dbAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_PROFAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SERVAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SITE

Only the SQL 2005 host instance and database name combination.

APPLICATIONS:\SqlServerWriter\host%5Cinstance\masterFor example, to back up the database named test11_SITE on the SQL Server named instance GALL, type the following:APPLICATIONS:\SqlServerWriter\GALL%5Ctest11_SITE

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 in SharePoint 2003 backup” on page 155 provides more information about URL encoding.

All SQL 2000 named instances and stand-alone databases.

APPLICATIONS:\MSDEWriter-SQL2000

Only the SQL 2000 host instance and database name combination.

APPLICATIONS:\MSDEWriter-SQL2000\host%5Cinstance\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

Some special characters in a SQL Server named instance or 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.

Table 26 SQL save set syntax for SharePoint 2003 (page 2 of 2)





URL encoding for SQL save sets in SharePoint 2003 backup

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:



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:

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




\ %5C ? %3F

/ %2F ] %5D

" %22 [ %5B

% %25 } %7D

# %23 { %7B

& %26 ^ %5E

< %3C ‘ %60

> %3E | %7C

SharePoint 2003 backup and recovery 155


SQL application information variable settings in SharePoint 2003 backup


Table 28 SQL variable settings for application information attribute




NSR_DATA_MOVER=valueapplication information attribute

Specifies the hostname of the NMM 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• The proxy client hostnameIf 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 SQL Client resource, 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.



Performing SharePoint 2003 backupsSharePoint 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 supported by NMM. Chapter 2, “Configuring a Scheduled Backup,” describes these 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 resourceThe 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:









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.

Performing SharePoint 2003 backups 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, specify the save set name:


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.


URL-encoded values are used to represent special characters such as the 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.



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.



Performing a SharePoint 2003 recoveryThe 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 2003To recover SQL Server 2000 stand-alone databases:

1. Stop the SQL Server services for the SQL Server 2000 stand-alone database instances.

2. From the navigation tree, expand the Applications folder and the 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 2003If 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.

Performing a SharePoint 2003 recovery 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.

2. From the navigation tree, expand the Applications folder and the SQLServerWriter folder.



5. After the recovery is complete, start the SQL Server services for the SQL 2005 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.



5. After the recovery is complete, start the SQL Server services for the SQL 2005 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.

2. From the navigation tree, expand the Applications folder and the MSDEWriter-SQL2000 folder.





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)



For application servers such as SQL servers, consider backing 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 dataAssign the appropriate snapshot policy and Client resource to each backup group.




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 the EMC VSS Provider.

NetWorker modules and 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.

Performing a SharePoint 2003 recovery 161


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.

Table 29 SQL Server backup and recovery best practices and considerations (page 2 of 2)




SharePoint 2003 scenario of backup and recovery processThe 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 databasesThe 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.

SharePoint 2003 scenario of backup and recovery process 163


Viewing the databases as they appear in SQL Enterprise Manager UIFigure 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 setsThe 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-SQL2000APPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServerAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5CReportServerTempDBAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5CSPS01_Config_dbAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_PROFAPPLICATIONS:\MSDEWriter-SQL2000\GALL%5Ctest11_SERVAPPLICATIONS:\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 UIThe 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.



Figure 18 NMM backups available for recovery

Deleting the new databasesThe 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 databasesThe 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.



Verifying database recovery through the SQL Enterprise ManagerThe 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 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



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 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 AdministrationIn 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



SharePoint 2007 disaster recoveryOne 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 planningThe following sections describe the steps you must take as part of your disaster recovery plan.

SharePoint Server backup listTable 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.

Additional requirements for successful SharePoint disaster recoveryThe following conditions and requirements must be met for a successful disaster recovery:

◆ 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.

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" databaseAPPLICATIONS:\Microsoft Office SharePoint Services\



◆ 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 recovery” on page 109.

• SharePoint farm as described in “Full recovery of a stand-alone SharePoint farm” on page 149.

Backing up SharePoint critical componentsThe 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 CacheThe .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__71e9bce111e9429c

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.


7


◆ NMM support for granular backup and recovery of SharePoint items.............. 174◆ Better performance for large SharePoint farms ...................................................... 175◆ Performing SharePoint 2007 granular backups ...................................................... 175◆ Configuring a SharePoint 2007 Client resource for granular backup.................. 176◆ SharePoint 2007 granular recovery........................................................................... 177◆ SharePoint recovery search........................................................................................ 179◆ Performing a granular recovery of SharePoint items ............................................ 182◆ SharePoint prerecovery options ................................................................................ 183◆ SharePoint recovery options...................................................................................... 185◆ Commands for SharePoint 2007 granular backup and recovery ......................... 186◆ Using administrator account other than that created during installation .......... 192

Microsoft SharePoint2007 Granular Backup

and Recovery

Microsoft SharePoint 2007 Granular Backup and Recovery 173

Microsoft SharePoint 2007 Granular Backup and Recovery

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 backupGranular 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

◆ Within the SharePoint UI, retrieve the lost document from the SharePoint recycle bin.



Better performance for large SharePoint farmsNMM 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 backupsA 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 backupTo configure a SharePoint 2007 Client resource for granular backup:










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, specify the save set name:

NMMOSS:/URL=<url>

The SharePoint Server save set syntax is described in “Specifying save sets for SharePoint granular backup” on page 186.







nsr_moss_save.exe


OPTIMIZED=false







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.

18. Click OK.

SharePoint 2007 granular recoveryYou 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.



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

SharePoint recovery search 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 NMM displaying Search tab for SharePoint 2007 recovery



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 itemTo 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.

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 Select the type of range, Before, Between, or After and then specify the date or date range. If not selected, this field is ignored.

Modification Date Select the type of range, Before, Between, or After and then specify the date or 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 withlist 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.

SharePoint recovery search 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 itemsTo 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.




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.



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 recoveryNMM uses special commands for SharePoint 2007 granular backup and recovery.

nsr_moss_save.exe commandThe 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 backupTable 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 SharePoint granular recovery save set syntax for SharePoint 2007 objects


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 backupThe 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 -fName Save Set SyntaxSharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569NMMOSS:/SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569

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.

Commands for SharePoint 2007 granular backup and recovery 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 -fName Save Set SyntaxSharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569NMMOSS:/SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569

• nsr_moss_save [-w]

Displays all web applications and save set syntax. For example:

nsr_moss_save -wAll Web Applications:Name Save Set SyntaxSharePoint - 80 NMMOSS:/SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569/SharePoint - 80@CentralAdmin NMMOSS:/SharePoint_Config_5e687d39-3cf4-4769-b88b-1d3ead776569/@CentralAdminSharePoint - 26061 NMMOSS:/SharePoint_Config_5e687d39-3cf4-4769-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 - 80Name Save Set Syntaxhttp://mj-hvmoss NMMOSS:/Url=http:/mj-hvmosshttp://mj-hvmoss/sites/Sunil NMMOSS:/Url=http://mj-hvmoss/sites/Sunilhttp://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/dsdAll Sites for Site-Collection: http://inenbeherad2c:48137/sites/dsdName Saveset Syntaxhttp://inenbeherad2c:48137/sites/dsd/subsite NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd/subsitehttp://inenbeherad2c:48137/sites/dsd NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsdhttp://inenbeherad2c:48137/sites/dsd/subsite/subsite NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd/subsite/subsite



• 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/dsdAll Lists for Web: http://inenbeherad2c:48137/sites/dsdName Saveset SyntaxLinks NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Links;Type=ListTasks NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Tasks;Type=ListList 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=ListForm Templates NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Form Templates;Type=ListReporting Metadata NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Reporting Metadata;Type=ListConverted Forms NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Converted Forms;Type=ListTeam Discussion NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Team Discussion;Type=ListCalendar NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Calendar;Type=ListMaster Page Gallery NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Master Page Gallery;Type=ListAnnouncements NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Announcements;Type=ListReporting Templates NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Reporting Templates;Type=ListWeb Part Gallery NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/Web Part Gallery;Type=ListUser Information List NMMOSS:/Url=http://inenbeherad2c:48137/sites/dsd;Name=/User Information List;Type=ListSite 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 settingsTable 33 on page 189 describes the variable that can be specified in the application information attribute of the Client resource.

Table 33 Application information variable settings


-A OPTIMIZED=value For granular backup use -A OPTIMIZED=false.The default value is false. This value is required.

true | false

For example, nsr_moss_save.exe –s <server> -c <client> -A OPTIMIZED=false



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 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 0(default) - No, 1 - Yes

halton-nonfatal 0(default) - No, 1 - Yes

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...]



Using administrator account other than that created during installationTo run SharePoint granular backup and recoveries in NMM using an administrator account other than that created during NMM installation, the following permissions are required:

◆ 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


◆ Exchange Server backup and recovery .................................................................... 194◆ Performing Exchange Server backups ..................................................................... 206◆ Configuring an Exchange Client resource............................................................... 208◆ Performing Exchange Server recovery..................................................................... 219◆ Recovery to storage groups ....................................................................................... 226◆ Microsoft Exchange Server 2007 directed recovery ............................................... 238◆ Microsoft Exchange Server 2007 disaster recovery ................................................ 241

Microsoft ExchangeServer Backup and

Recovery

Microsoft Exchange Server Backup and Recovery 193

Microsoft Exchange Server Backup and Recovery

Exchange Server backup and recovery This chapter supplements the overall NetWorker Module for Microsoft Applications (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 supportAccess 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 clientNMM 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.

The EMC Information Protection Software Compatibility Guide provides the most up-to-date lists of hardware, operating system, service pack, and application versions that are supported by the NMM client.

NMM and Exchange Server support for L18NMicrosoft requires Exchange Server 2003 Storage Group names to be in the same locale as the Microsoft Exchange Server.

Exchange Server backup and recovery typesNMM supports several types of backup and recovery:



◆ 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.

◆ Microsoft Exchange Server 2007

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

Exchange Server backup and recovery 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 credentialsIf Exchange Server 2007 is installed after NMM is installed, then NMM must be reinstalled after Exchange is installed. During NMM installation, add Exchange credentials.

The EMC NetWorker Module for Microsoft Applications Release 2.2 Installation Guide provides more information about initial setup of Exchange credentials during NMM installation.

Modifying Exchange Server 2007 credentialsNMM 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:

◆ 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 recoveryNMM 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.



◆ 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.

Displaying valid Exchange data save setsWhen 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.

2. From the command prompt, type the following command to list the valid application data save set names:

• If the Microsoft Exchange server is a standalone node or LCR active node, type:

nsrsnap_vss_save -?

Table 35 Exchange 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 ExchangeWriter\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.



• 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 Writer""APPLICATIONS:\Microsoft Exchange Writer\first_group""APPLICATIONS:\Microsoft Exchange 2007""APPLICATIONS:\Microsoft Exchange 2007\second_group"

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. 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

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 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.



\ %5C ? %3F

/ %2F ] %5D

" %22 [ %5B

% %25 } %7D

# %23 { %7B

& %26 ^ %5E

< %3C ‘ %60

> %3E | %7C



Exchange application information variable settings


Table 37 Exchange application information variable settings (page 1 of 2)






• The local host• The proxy client hostnameIf 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 you are setting up an Exchange Client resource, it is recommended that you use a proxy client. A proxy client requires hardware for transportable snapshots. 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 Exchange server. Related messages are logged on the Client resource host in both the Replication Manager log files and 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=6728An 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.



Serverless backupA proxy client is used in conjunction with instant snapshot backups to provide 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.


NSR_CHECK_JET_ERRORS=value

Specifies which JET database errors to check for in the event log before taking a snapshot.

Note: If errors are found, the snapshot will fail.

• 447 (Exchange Server 2007 only)• 448 (Exchange Server 2007 only)• 1018• 1019• 1022• all• noneValues can be separated with a comma, for example:NSR_CHECK_JET_ERRORS=1018,1019The default is all.

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.

The name of the Exchange Virtual Server. For example: NSR_VIRT_SERV=virtual_server_nameThis value is required in a CCR passive 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_INDEX_CLIENT=virtual_server_nameThis value is required in a CCR passive node backup.

Table 37 Exchange application information variable settings (page 2 of 2)




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 where the NMM client software is installed.

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.



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 timeWhen 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.

LAN

NetWorkerStorage

Node 7.3.xPrimary storage

(For example,CLARiiON or Symmetrix)

Traditionalstorage

S1 S2

S3S4

S5


Exchange ServerNetWorker Module forMicrosoft Applications

Proxy clientNetWorker Module forMicrosoft Applications

SAN

GEN-000749



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.

◆ 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.



Performing Exchange Server backupsNMM 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 backupsThe 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 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 7.4The 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 2.2NMM 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



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 backupsIncremental 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 resourceA Client resource specifies what to include in a snapshot of a NMM client. Client resources are associated with other backup resources such as groups and snapshot 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.

Configuring an Exchange Client resource

To create an Exchange Client resource for Exchange standalone, LCR, or SCC:




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



• 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 sets for application data” on page 198.

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.


URL-encoded values are used to represent special characters such as the backslash (\), and a table of values is provided in “URL encoding for Exchange save sets” on page 199.


Note: The backup group must be a snapshot group.





Configuring an Exchange Client resource 209






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 CCRThe 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 nodesIn 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:

◆ “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 nodeIn 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:




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:











15. In the Backup command attribute, type the backup command:.




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 nodeWhen 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.

To create an Exchange Client resource if backup is performed from an Exchange CCR passive node:

Note: “Configuring deduplication backup for a passive node in a CCR environment” on page 218 provides additional steps for enabling deduplication.




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.





















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.



Configuring backup for an Exchange CCR active node with a data moverIn 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 with a data mover:




4. In the Name attribute, type the <Exchange_Virtual_Server_hostname>





















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 moverWhen 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.

To create an Exchange Client resource if backup is performed from an Exchange CCR passive node with a data mover:




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.





















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>








19. Click OK.



Configuring deduplication backup for a passive node in a CCR environmentTo 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.



Performing Exchange Server recoveryAfter 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 limitationsThe 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.

Performing Exchange Server recovery 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 recoveriesThere 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 mailboxesThere 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.



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 optionsTo 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



• 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.



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 recoveryThis 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.



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 recoveryCertain 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 NMMThis 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.



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.



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:

◆ <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 groupRecovery 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.

Recovery to storage groups 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 databasesBefore 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!When you use the restore to an RSG feature, always delete the existing RSG and 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.



• 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” –AllowFileRestore:$true

Recovering to the RSGWhen 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!When you use the restore to an RSG feature, always delete the existing RSG and 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 NMM client software was configured for backup.


3. From the left pane, select Recover Session > Exchange Recover, and then select Database Recover. This displays the available Exchange database backups.





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 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.



Figure 38 Public folders cannot be recovered to an RSG


Setting a different Exchange Server for recoveryNMM 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 mailboxesAn 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.



• <username> is the name of the user being granted permissions.

Additional requirements for browsing RSG mailboxesThere 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 databaseOnce 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 recoveryItems 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.



Figure 39 Selected and partially selected items

Searching for an itemYou 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.



• 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 itemsOnce 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 groupRecovery 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 databasesBefore 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.



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



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” –AllowFileRestore:$true

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.





Microsoft Exchange Server 2007 directed recoveryIn 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.

Performing a directed recoveryTo 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.



Figure 41 Local client on the taskbar next to Client


Figure 42 on page 239 shows the Configuration Options dialog box.

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

Refresh button

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.



Microsoft Exchange Server 2007 disaster recoveryNMM 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.

Microsoft Exchange Server 2007 disaster recovery 241

http://technet.microsoft.com/en-us/library/bb332343.aspx


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.



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.



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:\






• 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


Recovery site (in the same network as the production setups):



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.



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.


9


◆ Data Protection Manager backup and recovery ..................................................... 248◆ Performing DPM Server backups ............................................................................. 255◆ Configuring a DPM Client resource ......................................................................... 256◆ Performing DPM Server recovery ............................................................................ 258

Microsoft DataProtection Manager

Backup

Microsoft Data Protection Manager Backup 247

Microsoft Data Protection Manager Backup

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

◆ Microsoft Exchange Server

◆ Microsoft SQL Server

◆ Microsoft Office SharePoint Server

◆ 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.



Figure 45 NMM protection of DPM

This chapter supplements the overall NetWorker Module for Microsoft Application 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.

The EMC Information Protection Software Compatibility Guide provides the most up-to-date lists of hardware, operating system, service pack, and application versions supported by the NMM client.

Microsoft Servers:Backup and Recovery

provided by DPMServer 2007

DPM Server 2007:Backup and Recoveryprovided by NetWorkerModule for MicrosoftApplications (NMM) client

GEN-000754

NetWorker Server7.3.3 or later:Provides services forNetWorker clientssuch as NMM

NetWorker Module forMicrosoft Applicationsprotects items within

DPM Server 2007:DPM Configuration Database

DPM Replicas

DPM protects items within Microsoft Servers:

Database filesStorage groups

VolumesTransaction logs

LAN

Data Protection Manager backup and recovery 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



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_saveset

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.



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:

Displaying valid DPM data save setsTo 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 node that is currently hosting the application server.

3. Press Enter.


APPLICATIONS:\Microsoft DPM\DPM database.APPLICATIONS:\Microsoft DPM\Volume C: on server mars.saturn.company.com

Table 39 DPM Server 2007 save set syntax


Configuration databaseThere will always be exactly one DPM configuration database.

APPLICATIONS:\Microsoft DPM\DPM database.

Protection group replica set, of a non-VSS datasourceThere 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 setThere 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.




Turning the MSDE writer on or off in the registryAlthough 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\Settings\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 setsWhen 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 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




DPM application information variable settings


Table 40 DPM application information variable settings




NSR_PS_SINGLE_SAVE_PER_SAVESET=value

Specifies to create a separate PowerSnap save set for each save set in the client entry.

yesNSR_PS_SINGLE_SAVE_PER_SAVESET=yes

NSR_PS_FIRST_SAVESET_SAVE=value

Specifies the save set that 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 nameFor example:NSR_PS_FIRST_SAVESET_SAVE=MY_DPM_CONFIG_DB

NSR_DPM_RETRY_WAIT=value

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 setting is 1.

A positive integerFor example, to specify a wait time of two minutes: nsr_dpm_retry_wait=2

NSR_DPM_RETRY_MAXIMUM=value

Specifies the total number of times—the first try plus retries—to attempt to start the backup.

A positive integerFor example, to specify of three attempts: nsr_dpm_retry_maximum=3



Performing DPM Server backupsNMM DPM backup provides:


◆ Full backup of DPM databases, logs, and storage groups, as well as administrator-selected storage groups.

Configuring DPM backups




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 resourceA Client resource specifies what to include in a snapshot of a 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 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:










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.comAPPLICATIONS:\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.




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:


14. In the Application information attribute, type the following variables and values, each on a separate line:

NSR_PS_SINGLE_SAVE_PER_SAVESET=yesNSR_PS_FIRST_SAVESET_SAVE=APPLICATIONS:\Microsoft DPM\DPM databaseNSR_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 recoveryWhen the files are backed up, administrators can the perform the following types of recovery:

◆ “Performing a granular recovery of DPM objects” on page 260

◆ “DPM disaster recovery” on page 267


10


◆ Performing a granular recovery of DPM objects .................................................... 260◆ Performing a granular directed recovery to a DPM-protected server ................ 264◆ DPM disaster recovery ............................................................................................... 267

Microsoft DataProtection Manager

Granular and DisasterRecovery

Microsoft Data Protection Manager Granular and Disaster Recovery 259

Microsoft Data Protection Manager Granular and Disaster Recovery

Performing a granular recovery of DPM objectsWhen 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.

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.

LAN

DPM-protectedservers: MicrosoftApplications

SharePointServer3

SQLServer4

FileServer1

ExchangeServer2

Directed Recovery:Granular file recoveriesto DPM-protectedservers, directly fromNetWorker Server,through NMM client UI onthe DPM-protected server.

NetWorker Server:Provides backupservices forNetWorker clientssuch as NMM

DPM Server 2007:NMM client providesbackup of DPMServer andDPM-protectedserver replicas.

GEN-000958


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.

Performing a granular recovery of DPM objects 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.











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.

Performing a granular recovery of DPM objects 263


Performing a granular directed recovery to a DPM-protected serverIn 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 scenarioThe following scenario describes how DPM directed granular recovery would work with a hypothetical FileServer1 as shown in Figure 51 on page 264.

Figure 51 Directed recovery of FileServer1

LAN

DPM-protectedservers: MicrosoftApplications

SharePointServer3

SQLServer4

FileServer1

ExchangeServer2

Directed Recovery:Granular file recoveriesto DPM-protectedservers, directly fromNetWorker Server,through NMM client UI onthe DPM-protected server.

NetWorker Server:Provides backupservices forNetWorker clientssuch as NMM

DPM Server 2007:NMM client providesbackup of DPMServer andDPM-protectedserver replicas.

GEN-000958



Directed recovery in DPM scenarioIn 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.


By default, the objects displayed in the navigation tree are from the most recent backup. To recover objects from a previous backup:






18. Click OK.




DPM disaster recoveryIn 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.

DPM disaster recovery 267


Figure 54 Recover submenu for DPM recovery


6. In the navigation tree, select the DPM replicas or databases to be recovered.





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.



13. 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.

14. Click OK.


DPM disaster recovery 269

11


◆ Hyper-V backup and recovery.................................................................................. 272◆ Performing Hyper-V backups ................................................................................... 284◆ Configuring a Hyper-V Client resource................................................................... 286◆ Performing Hyper-V recovery .................................................................................. 288◆ Hyper-V parent partition disaster recovery ............................................................ 292

Microsoft Hyper-VBackup and Recovery

Microsoft Hyper-V Backup and Recovery 271

Microsoft Hyper-V Backup and Recovery

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

This chapter supplements the overall NetWorker Module for Microsoft Application 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.



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 partitionsHyper-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:

◆ Microsoft Exchange Server

◆ Microsoft Office SharePoint Server

◆ Microsoft SQL Server

◆ Microsoft Data Protection Manager

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.

Hyper-V backup and recovery 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.

Figure 56 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.

Physical server withWindows Server 2008host operating system

and Hyper-V role

Virtual Machine VS-1Windows Server 2008 x64Microsoft Exchange Server 2007

Virtual Machine VS-2Windows Server 2003Microsoft Office SharePoint Server 2003

Virtual Machine VS-3Windows Server 2003Microsoft Data Protection Manager

Virtual Machine VS-4Windows Server 2003Microsoft SQL Server

Hyper-V configuration file

Configure Client resourceson parent partition to backup parent partition volumes

and system components

Configure Client resources on parent partition toback up Hyper-V child partitions and Hyper-V

configuration fileGEN-000959

Hyper-V virtual machines with child partitionoperating systems and applications

Configure Client resources withineach child partitions to back upapplications, volumes, andsystem components on thechild partition



The Microsoft website provides more details and the most up-to-date information about storage hardware supported by Hyper-V.

The EMC Information Protection Software Compatibility Guide provides the most up-to-date information about what software and hardware providers are supported for Hyper-V backup and recovery in NMM.

Figure 57 Hyper-V storage options

LUN1 LUN 1 -> Y:

FC or iSCSI

LUN 2

Disk1-> X:

Disk 2

VHD2 -> E:

LUN2 -> F:

LUN 3 -> G:

VHD1 -> C:Configuration 1

Disk 2 -> D:

Y:\V2\V2.VHD-> VHD2

\\FS\SHARE1\V3\V3.VHD-> VHD3

X:\V1\V1.VHD-> VHD1

SAN

File Server

LUN2

LUN3

S:SHARE1

\\FS\SHARE1

FC or iSCSI

CIFS/SMB/SMBv2

iSCSI

DASParent Partition Child Partition

VHD3 -> H:

Configuration 2

Configuration 3

Configuration 4

Configuration 5

Configuration 6

GEN-000962

passthrough

passthrough

Hyper-V Server

Table 41 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



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.



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 partitionWindows 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:“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



On the parent partitionThe Hyper-V role can coexist with other applications, such as: • Microsoft SQL Server• Microsoft Office SharePoint Server• Microsoft Exchange Server• Microsoft System Center Data

Protection Manager (DPM) • Microsoft Windows Server Cluster “NetWorker Module features” on page 22 lists the specific versions supported for each of these applications.

Configure application backups with the specific instructions for the application in:

“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“Performing SQL Server backups” on page 119“Performing SharePoint 2007 backups” on page 141“Performing Exchange Server backups” on page 206“Performing DPM Server backups” on page 255“Performing Windows Server Cluster backups” on page 298

Hyper-V on parent partitionHyper-V virtual machines and Initial Store configuration file

Complete tasks 1 through 4:“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

• 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 applicationsApplication data, such as: • Microsoft SQL Server• Microsoft Office SharePoint Server• Microsoft Exchange Server• Microsoft System Center Data

Protection Manager (DPM) • Microsoft Hyper-V• Microsoft Windows Server Cluster “NetWorker Module features” on page 22 lists the specific versions supported for each of these applications.

Configure application backups with the specific instructions for the application in:

“Performing SQL Server backups” on page 119“Performing SharePoint 2007 backups” on page 141“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 42 Backup tasks for Hyper-V (page 2 of 2)

Hyper-V parent partition Tasks



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.



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 locationsHyper-V has several restrictions on relocating and restoring to other locations:

◆ 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.

.

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.

Figure 58 NMM backup of Hyper-V components

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.

Parent Partition:Windows Server 2008OS with Hyper-V Role

VS-1Virtual MachineChild Partition

“APPLICATIONS:\Microsoft Hyper-V”Save set backs up Hyper-V components, which includeseach virtual machine and the Hyper-V configuration file

“APPLICATIONS:\Microsoft Hyper-V\VS-1”Save set to back up virtual machine “VS-1”

VS-2Virtual MachineChild Partition

“APPLICATIONS:\Microsoft Hyper-V\VS-2”Save set to back up virtual machine “VS-2”

Hyper-V Configuration File

GEN-000960

“APPLICATIONS:\Microsoft Hyper-V\Initial Store”Save set to back up configuration file “Initial Store”



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.

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:

APPLICATIONS:\Microsoft Hyper-VAPPLICATIONS:\Microsoft Hyper-V\Initial StoreAPPLICATIONS:\Microsoft Hyper-V\virtual_machine_name_1APPLICATIONS:\Microsoft 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.


Table 44 Hyper-V save set syntax


Hyper-V ManagerThe entire Hyper-V configuration: the Hyper-V configuration file, and each virtual machine

APPLICATIONS:\Microsoft Hyper-V

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.

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.

APPLICATIONS:\Microsoft Hyper-V\Initial Store

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. There are usually multiple virtual machines on the host operating system.

APPLICATIONS:\Microsoft Hyper-V\virtual_machine_nameVirtual machines can be included in a proxy backup group.



Hyper-V application information variable settings


Table 45 Hyper-V application information variable settings






Performing Hyper-V backupsNMM Hyper-V backup provides full backup of Hyper-V virtual machines and Hyper-V Initial Store configuration file.

Preparing a virtual machine for backupBefore 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_1APPLICATIONS:\Microsoft Hyper-V\virtual_machine_name_2



The Microsoft Hyper-V save set and configuration file (Initial Store) save set cannot be specified in a proxy backup group:

APPLICATIONS:\Microsoft Hyper-VAPPLICATIONS:\Microsoft Hyper-V\Initial Store

Configuring Hyper-V backups




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 resourceA Client resource specifies what to include in a snapshot of a 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 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:










To back up all of Hyper-V—the virtual machines and the configuration file (Initial Store)—specify:

APPLICATIONS:\Microsoft Hyper-V

To back up the virtual machines, specify:

APPLICATIONS:\Microsoft Hyper-V\virtual_machine_name_1

APPLICATIONS:\Microsoft Hyper-V\virtual_machine_name_2

To back up the configuration file (Initial Store), specify:

APPLICATIONS:\Microsoft Hyper-V\Initial Store

Note: The Microsoft Hyper-V save set and configuration file (Initial Store) save set cannot be specified in a proxy backup group:APPLICATIONS:\Microsoft Hyper-VAPPLICATIONS:\Microsoft Hyper-V\Initial Store



“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.


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 recoveryWhen 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.



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.

Performing Hyper-V recovery 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.



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.

Performing Hyper-V recovery 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.

Figure 59 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

◆ Hyper-V on host: Hyper-V virtual machines and Initial Store configuration file

◆ Hyper-V in-guest: Application data

Physical server withWindows Server 2008host operating system

Virtual Machine VS-1Windows Server 2008 x64Microsoft Exchange Server 2007

Virtual Machine VS-2Windows Server 2003Microsoft Office SharePoint Server 2003

Virtual Machine VS-3Windows Server 2003Microsoft Data Protection Manager

Virtual Machine VS-4Windows Server 2003Microsoft SQL Server

Hyper-V configuration file

1. On the physical serverhost, recover the hostvolumes and systemcomponents.2. Re-enable the Hyper-Vrole.

3. On the physical server host, recover theHyper-V virtual machine guests andHyper-V configuration file.

GEN-000961

Hyper-V virtual machines with guestoperating systems and applications

4. On each guest, recover theapplications using the recoveryprocess appropriate for eachapplication.



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 recovery” on page 288.

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

12


◆ Windows Server Cluster backup and recovery ...................................................... 296◆ Performing Windows Server Cluster backups........................................................ 298◆ Windows Server Cluster recovery ............................................................................ 299◆ Windows Server 2003 clustered client disaster recovery ...................................... 299

Microsoft WindowsCluster Backup and

Recovery

Microsoft Windows Cluster Backup and Recovery 295

Microsoft Windows Cluster Backup and Recovery

Windows Server Cluster backup and recovery This chapter supplements the overall NetWorker Module for Microsoft Application (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 recoveryNetWorker 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:

◆ 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





• The local host• The proxy client hostnameIf 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 Windows Cluster client resource, it is recommended that you use a proxy client. If a proxy client is 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.

If no proxy client is specified for an Windows Cluster 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.

Windows Server Cluster backup and recovery 297


Performing Windows Server Cluster backupsTo 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.

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 backupsIf 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 Navisphere user interface or the NaviCLI command line interface to search for and delete the inactive snapshot.

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.



Windows Server Cluster recoveryThe 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.

c. From the Options menu, select System Recover Configuration Options and click NetWorker.







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 settingsIf 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.

Windows Server 2003 clustered client disaster recovery 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 serverThis 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



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 controllerThis 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.

c. From the Options menu, select System Recover Configuration Options and click the NetWorker tab.







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.



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.


13


◆ Active Directory backup and recovery .................................................................... 308◆ Performing an Active Directory or ADAM backup............................................... 309◆ Performing a granular Active Directory or ADAM backup................................. 310◆ Recovering all Active Directory data ....................................................................... 318◆ Performing a granular recovery of Active Directory objects and attributes ...... 320◆ Active Directory disaster recovery ........................................................................... 324

Microsoft ActiveDirectory Backup and

Recovery

Microsoft Active Directory Backup and Recovery 307

Microsoft Active Directory Backup and Recovery

Active Directory backup and recoveryThis chapter supplements the overall NetWorker Module for Microsoft Applications (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.

Types of supported backup and recovery 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

◆ 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 backupPerforming 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 backupA 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 backupThe following are general guidelines to consider before performing a granular Active Directory backup:

◆ 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 5: Configure a Client resource” on page 315

◆ “Task 6: Configure privileges” on page 317

Task 1: Set up local Windows SYSTEM account as an ADAM administratorThe 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.

Performing a granular Active Directory or ADAM backup 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 deviceConfigure 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.



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 templateLabels 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:


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:


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 backups for this pool.

13. Click OK.

Labeling the deviceBefore a device can be used for backup, it must be labeled.

To label a device:

1. In the Administration window of the NetWorker Management Console, click 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.

Note: Level 1 to 9 backups are not supported for granular Active Directory and ADAM backups.

Task 4: Configure a backup groupBackup 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.

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\.



To configure a backup group:


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.

Task 5: Configure a Client resourceA 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:






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.





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



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:


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.



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.




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 attributesYou 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 NetWorker server on which the NetWorker client was configured for backup.

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.





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 restrictionWhen 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 recoveredObject 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

Performing a granular recovery of Active Directory objects and attributes 321


Attributes that are retained for a deleted objectTable 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.

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 attributesAn object’s password is not recovered. After recovering an object with a password attribute, the Windows administrator must reset the password.

Moved or renamed objectsIf 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.

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



Attributes with null valuesAttributes 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 objectsSchema objects cannot be recovered and therefore, should never be deleted.

Performing a granular recovery of Active Directory objects and attributes 323


Active Directory disaster recoveryA 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:

◆ 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


◆ Roll over a snapshot.................................................................................................... 326◆ Roll back a snapshot ................................................................................................... 327◆ Delete a snapshot ........................................................................................................ 330

SnapshotManagement

Snapshot Management 325

Snapshot Management

Roll over a snapshotA 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:

1. Open the NMM client software and select the NetWorker server on which the 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.


Snapshot Management

Roll back a snapshotA 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 requirementsTo 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 fileA 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.

Roll back a snapshot 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=2Storage_Array_1=APM00052106710

[APM00052106710]Array_Type=CLARiiONArray_model=CX500Array_user=sbarmaArray_pass=****Array_pass_confirm=****Array_SPA=122.222.111.128 Array_SPB=122.222.111.129


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=1Storage_Array_1=000187461273

[000187461273]Array_type=SymmetrixArray_sid=197Array_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.

Roll back a snapshot 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 snapshotSnapshot 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:

1. Open the NMM client software and select 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 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.


15


◆ Specifying ranges for service and connection ports............................................... 332◆ Specifying a password................................................................................................ 333

NetWorker ClientManagement

NetWorker Client Management 331

NetWorker Client Management

Specifying ranges for service and connection portsA 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.



Specifying a passwordThe 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

16Invisible Body Tag


◆ The troubleshooting process...................................................................................... 336◆ NMM client error messages....................................................................................... 338◆ NMM client issues....................................................................................................... 340◆ Checking log files ........................................................................................................ 347◆ Manually stopping and starting services................................................................. 350◆ Other troubleshooting resources............................................................................... 351

Troubleshooting

Troubleshooting 335

Troubleshooting

The troubleshooting processNetWorker 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.

Figure 60 Troubleshooting process

No

Yes

NetWorker Server

Unique toNMM Client

(volumes andwriters)?

NetWorkerClient-side orServer-side?

NetWorker Client

Yes

No

No

Yes

No

Yes

Find error

Does the errorappear in a log file

or messagewindow?

Check NMM Clienterror messages

section

Try adifferentsolution?

Fixed? Finish

ContactTechnicalSupport

Check other EMCdocumentation

Check NMMClient issues

section

GEN-000753


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 ProviderThe logging level of the log file is controlled by the debug level of the process.

The troubleshooting process 337

Troubleshooting

NMM client error messagesError 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_SUPPORTEDThis 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.


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 deniedThis 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 issuesThe 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 backupA 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 failsIf a node within a cluster undergoes failover during a backup operation, the operation will not be successful. The next scheduled backup operation will be the next valid backup.

Solution

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 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

The EMC NetWorker Module for Microsoft Applications Release 2.2 Installation Guide 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.


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 serviceis 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 arrayThis 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."


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 ExchangeClient 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 ServerClient 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."


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.XMLSQL Writer.XMLetc.

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.xmlC:\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 passwordIf 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.


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 not be able to display backups.

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.


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.


Troubleshooting

Checking log filesThere 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 interface outputs for NMM client.

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 interface outputs for NMM client.


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 servicesNMM 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 Services used in NMM client

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.

winworkr NetWorker Process Provides graphical user interface for NetWorker software.


Troubleshooting

Other troubleshooting resourcesDoes 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


Glossary

This glossary contains terms related to disk storage subsystems. Many of these terms are used in this manual.

Aad hoc backup See manual backup.

administrators group 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 (applicationspecific 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 Volume used to store archive data. Archived data cannot be stored on a backup volume or a clone volume.

auto mediamanagement

Feature that enables the storage device to automatically label, mount, and overwrite an unlabeled or recyclable volume.

autochanger See library.

autochanger sharing See library sharing.

Bbackup Operation that saves data to a volume.

See also conventional backup and snapshot.

backup componentsmetadata 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.


Glossary

bootstrap 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.

browse policy 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.

Ccarousel See library.

client Computer, workstation, or fileserver whose data can be backed up and recovered.

client file index 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.

client-initiatedbackup

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 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.

cluster 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 Line on a display screen, also known as a command prompt or shell prompt, where you type software commands.

component 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.

componentsmetadata document

See metadata document.

consistent State of a dataset that is fully and immediately available to an application view.

Console server 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.

Ddomain controller Computer that stores directory data and manages user interactions within a domain,

including logon, authentication, directory searches, and access to shared resources.

data mover (DM) Client system or application, such as NetWorker, that moves the data during a backup, recovery, or snapshot operation. See also proxy client.

data retention policy See retention policy.

datawheel See library.

datazone Group of hosts administered by a NetWorker server.

Dynamic DriveSharing (DDS)

Feature that allows NetWorker software to recognize shared drives.

device 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 FileSystem)

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 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.

disk subsystem 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.

Ffile index See client file index.

file system 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.

Ggroup Client or group of client computers that are configured to back up files at a

designated time of day.


Glossary

Hhigh-available system 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.

host ID Serial number that uniquely identifies a host computer.

Iinactivity timeout Number of minutes to wait before a client is considered to be unavailable for backup.

incremental backup Backup level in which only files that have changed since the last backup are backed up. See also level.

instant backup 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.

instant restore Process of copying data created during an instant backup to its original location, or to an alternate location, during a recover operation.

Jjukebox See library.

label Electronic header on a volume used for identification by NetWorker or other data mover application.

legacy method Use of special-case Microsoft APIs to back up and recover operating system components, services, and applications.

level 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.

library 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.

library sharing Shared access of servers and storage nodes to the individual tape drives within a library.

local cluster client 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.

locale settings Settings that specify the input and output formats for date and time, based on local language conventions.

LUN (logical unit) 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.

Mmanual 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 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.

media index 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.

NNDMP (Network Data

ManagementProtocol)

TCP/IP-based protocol that specifies how heterogeneous network components communicate for the purposes of backup and recovery.

NetWorkeradministrator

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.

NetWorker client See client.

NetWorker Consoleserver

See Console server.

NetWorkerManagement

Console

See Console server.

NetWorker server 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.

NetWorker storagenode

See storage node.

nonclone pool Pools that contain data that has not been cloned.


Glossary

nonpersistentsnapshot

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.

Oonline indexes Databases located on the NetWorker server that contain all the information

pertaining to the client backups (client file index) and backup volumes (media database).

operator Person who monitors the server status, loads backup volumes into storage devices, and executes day-to-day NetWorker tasks.

Ppathname 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.

persistent snapshot Snapshot that is retained on disk. A persistent snapshot may or may not be rolled over to tape.

PiT (point-in-timecopy)

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 Feature to sort backup data to selected volumes.

PowerSnap 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.

provider 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.

proxy client 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.

Qquiescing 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.

Rrecover To recover files from a backup volume to a client disk.


Glossary

Registry Microsoft Windows database that centralizes all Windows settings and provides security and control over system, security, and user account settings.

requestor Interface with the Microsoft VSS infrastructure to initiate the creation and destruction of shadow copy. NetWorker software is a requestor.

replica See shadow copy.

resource 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 To locate and recover archived files and directories.

rollback restore 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.

rollover 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.

root Highest level of the system directory structure.

Ssave set Group of files or a file system from a single client computer, which is backed up on

storage media.

SSID (save set ID) Internal identification number assigned to a save set.

save set recover To recover data by specifying save sets rather than by browsing and selecting files or directories.

save set status 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.

server index See client file index.

serverless backup 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).


Glossary

shadow copy set 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.

shadow copytechnology

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.

skip 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 Point in time, read-only copy of data created during an instant backup.

snapshot expirationpolicy

Policy that determines how long snapshots are retained before their storage space is made available for the creation of a new snapshot.

snapshot policy 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 retentionpolicy

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 See device.

storage node Storage device physically attached to a computer other than the NetWorker server, whose backup operations are administered from the controlling NetWorker server.

Vvolume 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 Internal identification that NetWorker software assigns to a backup volume.

volume mount point 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.

volume name Name assigned to a backup volume when it is labeled. See also label.


Glossary

volume pool See pool.

Volume Shadow CopyService (VSS)

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.

VSS See Volume Shadow Copy Service (VSS).

VSS component Subordinate unit of a writer.

Wwriter 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.


Glossary


Index

AActive Directory (AD)

backups for disaster recovery 309backups for object selection granularity 310disaster backup and recovery 38granular backup and recovery 38log files 349

Active Directory Lightweight Directory Services (AD LDS) 93, 308

active node 222AD LDS backup and recovery 308AD recovery

all objects for a disaster recovery 318granular recovery 320individual objects and attributes 320null value restriction 323restrictions for 321schema object restriction 323system-only attribute restriction 321tombstone restriction 321

ADAMauthoritative restore 94nonauthoritative restore 94recovery options 93

AES encryptionconflict with EFS encryption 100

All save setconsiderations for 84

alternate locationSharePoint directed recovery 183

alternate mount path for proxy client 200alternate storage group, Exchange recovery to 235application data save sets

considerations for 51, 122considerations for SharePoint 2003 161

application information attributeapplications 65Cluster Writer 297DPM 254Exchange 200Hyper-V 283SharePoint 2003 SQL data 156SharePoint 2007 139

SQL 118Application Specific Module (ASM) pass phrase 100application writers

DPM 251Exchange 197Hyper-V 278, 281SharePoint 2003 153SharePoint 2007 136supported 31

authoritative restoreADAM 94Cluster Writer 299DFS 95, 96FRS 97NTDS 99

Avamar 73configuration 75

Bbackup

Active Directory 310LCR 195SharePoint 2003 152

backup and recoveryHyper-V parent partition 273

backup client resourceconfiguring 61, 315configuring DPM 256configuring Hyper-V 286configuring multiple 61, 256, 286, 315

backup command attributeapplications 85DPM 257Exchange 210, 211, 213, 214, 216Hyper-V 287SharePoint 2007 144, 146, 147SharePoint 2007 granular 176

backup groupsconfiguring 59, 314

backup levels,deduplication 76

backup poolcreating for snapshots 52, 312


Index

backup roadmapHyper-V 274

backup save setsapplications 61DPM 252Exchange 198Hyper-V 282, 286SharePoint 2003 SQL data 153SharePoint 2007 137SharePoint 2007 disaster recovery 170specifying 84, 208, 211, 213, 214, 216, 256, 316SQL 115

backup scheduleconfiguring 58, 314

backup timeselecting for recovery 45

backupsdeduplication 76

best practicesSQL backup and recovery 121

best practices for backup and recoveryExchange 203Hyper-V child partition 279Hyper-V parent partition 279SharePoint 2003 161SQL in SharePoint 2003 161

browse timeselecting for recovery 45

Cchild partition

best practices for backup and recovery 279Hyper-V configurations 274

child partitions 273CLARiiON

E_VETO_PROVIDER error message 338error message 338provider overview 32replica takes long time 342

client resourceconfiguring for backup 61, 315configuring for deduplication 78configuring for DPM backup 256configuring for Exchange backup 206configuring for Hyper-V backup 286configuring for SharePoint 2003 backup 157configuring for SharePoint 2007 backup 143configuring for SQL backup 119configuring multiple 61, 256, 286, 315configuring when in a cluster 69, 298

Cluster Continuous Replication (CCR)Exchange backup 195

cluster serverentering physical hostname of 85, 257, 317scheduling backups for 69, 298

Cluster tab 299Cluster Writer recovery

authoritative restore 299for Windows Server 2008 91nonauthoritative restore 299

options 91common provider

See EMC VSS Provider 29component system writers

supported 31configuration database

backup 135SharePoint 2007 129SharePoint 2007 database backup 147SharePoint 2007 recovery 149

connection port rangespecifying 332

connection to NetWorker server 43consistency checker

utility 68conventional recovery

overview 36conventional snapshot policy example 57

Ddata deduplication 72deduplication

Avamar integration 73backup configuration notes 77backups 76configuration 75configuring client 78prerequisites 73querying deduplication save sets with mminfo 80support 74unsupported 74

Deduplication Nodeconfiguration 75

deduplication, data 72deleting snapshots 330device

configuring for rollover backup 52, 312DFS recovery options 94directed recovery

DPM-protected server 264Exchange Server 2007 238NMM Client 105SharePoint 2007 restore to an alternate location 183system 105

directory services restore modebooting in for FRS recovery 102booting in for NTDS recovery 102

disaster recoverybackup 37clustered client from a legacy non-VSS backup 304DPM 267Hyper-V 292Hyper-V parent partition 292legacy VSS backup 302Microsoft Exchange Server 2007 241NMM client 109NMM Client not on a domain controller 110NMM Client on a domain recovery 111planning for active directory 309recovery for active directory 318


Index

SharePoint 2007 170SQL 123support in NMM 38SYSTEM COMPONENTS 109Windows Server 2003 clustered client 299

DNS resolution, NetWorker Deduplication Node 75DPM

configuring client resource for backup 256directed recovery 264disaster backup and recovery 38disaster recovery options 269granular backup and recovery 38granular recovery 260URL encoding for save sets 253

DPM disaster recoveryindividual objects and attributes 267

EEFS encryption

conflict with AES encryption 100email notification

NetWorker events 337EMC VSS Provider

overview 29encryption

recovering backup data 86errors

E_VETO_PROVIDER 338E_VOLUME_NOT_SUPPORTED 338error code 0x800423f3 338Microsoft I/O write 342NMM client 338replica in a CLARiiON array 342VSS CLIENT... Invalid Writer Selection 339VSS_E_VOLUME_NOT_SUPPORTED_BY_PROVIDE

R 341VSS_E_WRITERERROR_RETRYABLE 338with Exchange 339

Event viewer 337events

notification 340exceptions

Windows Firewall 340Exchange

client resource configuration 206consistency checker utility 68disaster backup and recovery 38errors 339granular backup and recovery 38point-in-time recovery 223proxy client 201recover to Recovery Storage Group (RSG) 226roll-forward recovery 222URL encoding for save sets 199

Exchange backupCCR environment 195LCR environment 195

Exchange recoveryalternate storage group 235database can be overwritten selection 219

options for 219original storage group 227Remote Storage Group (RSG) 226

Exchange Server 2007directed recovery 238

exclusion representation of save sets 40

Ffailover from the passive node 222file system

disaster backup and recovery 38granular backup and recovery 38recovery options 89, 91

FRS recoverybooting in directory services restore mode 102consideration for 87options 97

full backup levelconsideration for 58, 314

Ggranular backup

SharePoint 2007 175, 176granular backup and recovery

support in NMM 38granular recovery

Active Directory 320backup 37DPM objects 260SharePoint 2007 177

groupsconfiguring for backup 59, 314

guest. See also child partition

Hhardware providers 32host

Hyper-V 274Host. See also Hyper-V parent partitionHyper-V

backing up a virtual machine with multiple volumes 284

backup roadmap 274components used for backup and recovery 281configuring backups 285configuring client resource for backup 286disaster backup and recovery 38disaster recovery 292granular backup and recovery 38, 280host 274parent partition disaster recovery 292recovery 288save sets 281VSS writer 281

Hyper-V best practiceschild partition backup and recovery 279parent partition backup and recovery 279

Hyper-V child partitionconfigurations 274


Index

Hyper-V parent partitionbackup and recovery types 273

Hyper-V storage and backup options 274

Iincremental backup level

consideration for 58, 314instant backup

overview 34instant recovery

overview 36instant snapshot policy example 57interface

overview 39issues

NMM client 340items

searching for recovery 44, 234

Llabel

creating for pool 53, 313large SharePoint farms 175LCR 195

Exchange backup 195LCR backup support 195level 1 to 9 backups

consideration for 58, 314level for backup

considerations for 58, 314Local Continuous Replication (LCR) 195log files 337

Active Directory 349NetWorker Server 349, 350NMM client 347PowerSnap client 347Replication Manager 348Solutions Enabler 348troubleshooting with 347

Mmarking items for recovery 44, 233metadata recovery 174Microsoft Cluster Service 296Microsoft Exchange Server 2007

disaster recovery 241Microsoft VSS software provider

overview 29Microsoft Windows Server 2008 31, 62, 91, 93, 250, 253,

296, 308Microsoft Windows Server 2008 support 91mminfo, querying deduplication save sets using 80monitoring a recovery 86, 102, 305mount path for proxy client 200MSCS 296MSDE

turning writer on or off in registry 253multiple client resources for backup 61, 256, 286, 315

NNetWorker Deduplication Node, DNS resolution 75NetWorker events

email notification 337NetWorker Monitor window 337Networker recovery options 89, 91NetWorker Server

log files 349, 350NetWorker server

connecting to 43NMM Client

directed recovery 105NMM client

error messages 338issues 340log files 347

nonauthoritative restoreADAM 94Cluster Writer 299DFS 95, 96FRS 97NTDS 99

nonpersistent backupoverview 34

notificationNetWorker events 340savegroup failed 340

notifications 337types 337

NSR_ALT_PATH command 200NSR_DATA_MOVER command 65NSR_DM_PORT command 200nsr_moss_save 186nsr_sa_ini.exe 327NTDS recovery

booting in directory services restore mode 102consideration for 87options 98

NUMBER_OF_PROCESSOR 175

Ooverview

conventional recovery 36Exchange proxy client 201instant backup 34instant recovery 36interface 39nonpersistent backup 34persistent backup 34provider 29proxy client 34, 118requestor 29rollback recovery 36rollover backup 34serverless backup 34, 118, 201software provider 29VSS writers 30

overwrite prevention errorExchange database recovery 342


Index

Pparent partition

best practices for backup and recovery 279pass phrase

recovering AES encrypted data 100passive node 222, 223password

changing 333consideration for changing 333restriction for active directory recovery 322specifying for backup 333

password-protectedrecovering backup data 86

persistent backupoverview 34

physical serverHyper-V 274

policies for snapshots 55pool

creating for snapshots 52, 312label for 53, 313

portsExchange proxy client communications 200service and connection port ranges 332

PowerSnap clientlog files 347

prerequisitesdeduplication 73

provideroverview 29

proxy clientconfiguring 67mount path 200overview 34, 118port number for communicating 200requirements for 67SharePoint serverless backup 139specifying hostname of 65SQL serverless backup 118SQL serverless backup in SharePoint 2003 156

Rrecover.exe 86recovery 179

ADAM options 93browse time selection 45Cluster options 91Cluster Writer options 91default method, choosing 90DFS options 94encrypted backup data 86Exchange options 219FRS options 97Hyper-V 288marking items 44, 233NTDS options 98password-protected backup data 86performing for system data 87searching for items 44, 234selecting a version 45

SharePoint 2007 granular 173SharePoint 2007 restore to an alternate location 183snapshots that have not been rolled over 101viewing progress for 86, 102, 305volumes required for 46

recovery errorExchange database 342

Recovery Storage Group (RSG)Exchange 226

recovery, deduplication 80remote access attribute

virtual cluster server specification 85, 257, 317replica

error message 338takes long time in CLARiiON environment 342

Replication Managererror message 338log files 348

requestoroverview 29

resource for backupconfiguring 61, 315configuring multiple 61, 256, 286, 315

rollback recoverylimitations of 327overview 36performing 327, 329requirements for 327

rollover backupdevice for 52, 312overview 34performing 326

Ssa.ini 327, 328safe mode, booting in 102save sets

applications 61DPM 252Exchange 198, 208, 211, 213, 214, 216exclusion representation 40Hyper-V 281, 282SharePoint 2003 154SharePoint 2003 SQL data 153SharePoint 2007 137SharePoint 2007 granular backup 186specifying for backup 84, 256, 316specifying for Hyper-V backup 286SQL 115SQL databases in SharePoint 2007 170URL encoding 63

savegroup failed notification 340Search Index

backup in SharePoint 2007 137searching for a recovery item 44, 234Security Configuration Wizard 340selecting items for recovery 44, 233server

connecting to NetWorker 43serverless backup 32, 118, 156


Index

Exchange 201overview 34proxy client requirement 67SharePoint 2007 backup 139SQL 118

service port rangespecifying 332

services 350SharePoint 2003

application save sets 161backup and recovery process 163backup and recovery support 152backups 157best practices for backup and recovery 161Client resource 157configuring Client resource 157recovery 159save sets 154SQL 2005 database recovery 159SQL Server 2000 recovery 159

SharePoint 2007configuration database 129configuration database backup 135configuration database recovery 149directed recovery through restore to an alternate

location 183disaster backup and recovery 38granular backup and recovery 38granular backup save sets 186granular backup storage requirements 133granular recovery 177granular recovery prerecovery options 182save sets for granular backup 176Search Index 135serverless backup 139specifying save sets for backup 143, 145, 147URL encoding 138URL encoding for save sets 138

snapshot policiesconfiguring 55examples of 57

snapshotsdeleting 330pool for 52, 312recovering when not rolled over 101rollback operation 327rollover operation 326

software provideroverview 29

Solutions Enablerlog files 348

SQLdatabase recovery for SharePoint 2003 159disaster backup and recovery 38serverless backup 118serverless backup in SharePoint 2003 156URL encoding 117URL encoding for SharePoint 2003 155

SQL proxy client 118SQL proxy client in SharePoint 2003 backup 156

SQL Serverconfigure client resource 119

storage and backup optionsHyper-V 274

storage arraycreating descriptor 328descriptor file 327Exchange rollback requirements 225

storage requirementsSharePoint 2007 granular backup 133

Symmetrixprovider overview 32

syntaxapplication save sets 61DPM save sets 252Exchange save sets 198Hyper-V save sets 282SharePoint 2003 SQL data 153SharePoint2007 137SQL 115

systemdirected recovery 105

system component writerssupported 31

SYSTEM COMPONENTSdisaster recovery 109

system recoveryperforming 87

Ttime of backup

selecting for recovery 45troubleshooting

checking other documents 351

UURL encoding

DPM save sets 253Exchange save sets 199save sets 63SharePoint 2007 138SharePoint 2007 save sets 138SQL save sets 117SQL save sets for SharePoint 2003 155SQL save sets in SharePoint 2003 backup 155

Vversion

selecting for recovery 45virtual cluster server

entering physical hostname of 85, 257, 317scheduling backups for 69, 298

Volume Shadow Copy Service (VSS)writers 30

volumes required for recovery 46VSS CLIENT... Invalid Writer Selection 339VSS_E_VOLUME_NOT_SUPPORTED_BY_PROVIDER

341VSS_E_WRITERERROR_RETRYABLE 338


Index

WWindows Encrypting File System (EFS) 100Windows Firewall

NetWorker blocked by 340Windows Server 2003

disaster recovery 299Windows Server 2008 31, 62, 91, 93, 250, 253, 308Windows Server 2008 Cluster Writer

nonauthoritative restore 91Windows Server Failover Clustering (WSFC) 296Windows SharePoint Services 2.0

see SharePoint 2003 152Windows SharePoint Services 3.0

see SharePoint 2007 128Windows system component writers

supported 31


Index
