HPE Data Protector 9.08 Integration guide for Microsoft Volume

HPE Data Protector 9.08 Integration guide for Microsoft Volume
HPE Data Protector
Software Version: 9.08
Integration guide for Microsoft Volume Shadow Copy
Service
Document Release Date: October 2016
Software Release Date: October 2016
Integration guide for Microsoft Volume Shadow Copy Service
Legal Notices
Warranty
The only warranties for Hewlett Packard Enterprise Development LP products and services are set forth in
the express warranty statements accompanying such products and services. Nothing herein should be
construed as constituting an additional warranty. HPE shall not be liable for technical or editorial errors or
omissions contained herein.
The information contained herein is subject to change without notice.
Restricted Rights Legend
Confidential computer software. Valid license from HPE required for possession, use or copying. Consistent
with FAR 12.211 and 12.212, Commercial Computer Software, Computer Software Documentation, and
Technical Data for Commercial Items are licensed to the U.S. Government under vendor's standard
commercial license.
Copyright Notice
© Copyright 2016 Hewlett Packard Enterprise Development LP
Trademark Notices
Adobe™ is a trademark of Adobe Systems Incorporated.
Microsoft® and Windows® are U.S. registered trademarks of Microsoft Corporation.
UNIX® is a registered trademark of The Open Group.
This product includes an interface of the 'zlib' general purpose compression library, which is Copyright ©
1995-2002 Jean-loup Gailly and Mark Adler.
Documentation Updates
The title page of this document contains the following identifying information:
l
l
l
Software Version number, which indicates the software version.
Document Release Date, which changes each time the document is updated.
Software Release Date, which indicates the release date of this version of the software.
To check for recent software updates: https://softwaresupport.hpe.com/patches.
To verify that you are using the most recent edition of a document, go to:
https://softwaresupport.hpe.com/manuals.
This site requires that you register for an HPE Passport and sign in. To register for an HPE Passport ID, go
to: https://hpp12.passport.hpe.com/hppcf/login.do.
Or click the Register link at the top of the HPE Software Support page.
You will also receive updated or new editions if you subscribe to the appropriate product support service.
Contact your HPE sales representative for details.
Support
Visit the HPE Software Support Online web site at: https://softwaresupport.hpe.com
This web site provides contact information and details about the products, services, and support that HPE
Software offers.
HPE Software online support provides customer self-solve capabilities. It provides a fast and efficient way to
access interactive technical support tools needed to manage your business. As a valued support customer,
you can benefit by using the support web site to:
HPE Data Protector (9.08)
Page 2 of 119
Integration guide for Microsoft Volume Shadow Copy Service
l
l
l
l
l
l
l
l
Search for knowledge documents of interest
Submit and track support cases and enhancement requests
Download software patches
Manage support contracts
Look up HPE support contacts
Review information about available services
Enter into discussions with other software customers
Research and register for software training
Most of the support areas require that you register as an HPE Passport user and sign in. Many also require a
support contract.
To register for an HPE Passport ID, go to:
https://hpp12.passport.hpe.com/hppcf/createuser.do
To find more information about access levels, go to:
https://softwaresupport.hpe.com/web/softwaresupport/access-levels
HPE Software Solutions Now accesses the HPESW Solution and Integration Portal Web site. This site
enables you to explore HPE Product Solutions to meet your business needs, includes a full list of
Integrations between HPE Products, as well as a listing of ITIL Processes. The URL for this Web site is
https://softwaresupport.hpe.com.
HPE Data Protector (9.08)
Page 3 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Contents
Chapter 1: Introduction
9
Chapter 2: Integration concepts
10
Volume Shadow Copy Service
10
Virtual Disk Service
12
VSSBAR agent
12
Backup types
13
VSS database
Querying the VSSDB
Deleting backup sessions
Enabling and disabling replicas
Changes in the environment
14
15
15
16
16
Chapter 3: Configuration
17
Prerequisites and limitations
Prerequisites
Transportable backup prerequisites
Limitations
ZDB and instant recovery limitations
HPE P6000 EVA Disk Array Family hardware provider limitations
HPE P9000 XP Disk Array Family hardware provider limitations
17
17
17
18
18
18
19
Configuring the application system for instant recovery-enabled backup sessions
19
Configuring HPE P4000 SAN Solutions
Prerequisites
Configuring the integration
Configuring the VSS hardware provider
20
20
21
22
Configuring HPE P6000 EVA Disk Array Family
Prerequisites
Configuring the integration
Considerations
Configuring the VSS hardware provider
22
22
23
23
23
Configuring HPE P9000 XP Disk Array Family
Prerequisites
Configuring the VSS hardware provider
Configuring the user authentication data
24
24
25
25
Configuring HPE 3PAR StoreServ Storage
Prerequisites
25
25
HPE Data Protector (9.08)
Page 4 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Configuring the Data Protector HPE 3PAR StoreServ Storage integration
Configuring the HPE 3PAR VSS hardware provider
26
26
Chapter 4: Backup
27
Backup types
27
Backup flow
27
Zero downtime backup with the Data Protector VSS integration
HPE P4000 SAN Solutions
HPE P6000 EVA Disk Array Family
HPE P9000 XP Disk Array Family
HPE 3PAR StoreServ Storage
28
28
28
29
29
Considerations
Microsoft Cluster environments and transportable backups
HPE P4000 SAN Solutions
HPE P6000 EVA Disk Array Family
Replica creation and reuse
HPE P9000 XP Disk Array Family
HPE 3PAR StoreServ Storage
Configuration check
30
30
30
31
31
31
32
32
Creating backup specifications
Backup options
ZDB options
33
37
38
Scheduling backup sessions
40
Starting backup sessions
42
Chapter 5: Restore
44
Standard restore
Restore modes
Component restore
File restore
Restoring using the GUI
Restore options
Restoring using the CLI
44
44
44
44
45
46
47
Instant recovery
Instant recovery methods
Switch of disks
Copy of replica data
Copy of replica data with the source volume retained
Copy of replica data with the source volume not retained
Restore snapshot data to the source volume
Restore snapshot data to the source volume with the source volume retained
Restore snapshot data to the source location with the source volume not retained
48
49
50
50
50
51
51
51
52
HPE Data Protector (9.08)
Page 5 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Resync of volumes
Limitations
HPE P4000 SAN Solutions considerations
HPE P6000 EVA Array considerations
HPE P9000 XP Array considerations
HPE 3PAR StoreServ Storage considerations
Instant recovery procedure
Limitations
Procedure
Chapter 6: Writer specifics
52
52
53
53
54
56
56
56
56
62
Microsoft Data Protection Manager writer specifics
Backup
Restore
Restore the DPM server first
Restore the DPM clients directly
65
65
67
67
68
Microsoft Exchange Server 2003 writer specifics
Backup
Restore
Rollforward recovery from the loss of one or more databases
Point-in-time restore after loss of a log file
Limitations
Instant recovery
Prerequisites
Post-instant recovery steps
Database recovery
User scenario for Microsoft Exchange Server 2003 backup and restore
Example - VSS transportable backup
Example restore scenario for Microsoft Exchange Server 2003
69
69
70
70
71
72
72
73
73
73
73
74
76
Microsoft Exchange Server 2007 writer specifics
Concepts
Continuous replication
Restore to original or another location
Backup
LCR and CCR environments
Restore
Standard restore
Rollforward recovery from the loss of one or more databases
Point-in-time restore after loss of a log file
Restoring individual mailboxes
Restoring a LCR or CCR copy to the original location
Instant recovery
Prerequisites
Limitations
Restoring an LCR or CCR copy to the original location
76
76
77
77
77
79
81
81
81
82
83
86
88
89
89
91
HPE Data Protector (9.08)
Page 6 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Post–instant recovery steps
Troubleshooting
91
91
Microsoft Exchange Server 2010 writer specifics
Introduction
Microsoft Exchange Server 2010 concepts
Integrating Data Protector Microsoft Volume Shadow Copy Service integration with
Exchange Server 2010
Configuration
Prerequisites
Licensing
Configuring
Backup
Limitations
Creating a backup specification
Restore
Restore scenarios
Restoring a passive copy
Restoring an active copy
Point-in-time restore
Instant recovery
Limitations
Instant recovery of a P6000 EVA Array snapclone to a different client using the CLI
Troubleshooting
92
93
93
93
93
93
94
94
94
94
94
97
98
99
99
99
100
Microsoft Hyper-V writer specifics
Concepts
Prerequisites
Backup
Prerequisites
Limitations
Backup from a physical cluster node
Restore
Restore from a physical cluster node
Troubleshooting
100
100
101
101
102
102
103
103
103
104
Microsoft SharePoint Services writer specifics
Concepts
Backup
Restore
104
104
105
106
MSDE writer specifics
Restore
108
108
Chapter 7: Troubleshooting
92
92
92
109
Before you begin
109
Checks and verifications
109
HPE Data Protector (9.08)
Page 7 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Problems
ZDB related problems
Send Documentation Feedback
HPE Data Protector (9.08)
109
114
119
Page 8 of 119
Chapter 1: Introduction
A traditional backup process is based on the direct communication between the backup application and the
application whose data is backed up. This backup method requires from the backup application an individual
interface for each application it backs up. To address the growing number of applications and their specific
interfaces, a new coordinator between the actors of the backup and restore process was introduced on
Microsoft Windows systems — the Microsoft Volume Shadow Copy Service.
Data Protector integrates with the Volume Shadow Copy service through the Data Protector Microsoft
Volume Shadow Copy Service integration (VSS integration).
The Data Protector integration with the Microsoft Volume Shadow Copy Service can serve in two functions:
l
l
It provides support for certified VSS writers . For a complete list of supported VSS writers and providers,
see the latest support matrices at https://softwaresupport.hpe.com/.
It forms the base for other integrations , such as the Data Protector Microsoft Exchange Server 2010
integration.
This guide describes how to use the generic Data Protector VSS integration for backing up and restoring
writer's data. For descriptions of application integrations, based on the VSS technology, see the appropriate
guides.
Benefits of using the VSS integration
Advantages of using the Data Protector VSS integration are the following:
l
l
l
A unified backup interface is provided for all applications that provide a writer.
Data integrity is provided on application level, because it is provided by the writers. No interference is
needed from the backup application.
Due to the generic nature of the programming and user interfaces, support for new applications, application
versions, or disk arrays (for example through certifications and so on) can be added much easier, even
after the initial Data Protector release.
Application integrations, such as the Data Protector Exchange Server 2010 integration, on the other hand
offer more functionality and are better adapted to the application specifics than the generic VSS
integration. However, due to additional implementation specifics, it may take longer to develop such
integrations as compared to adding support through the generic Data Protector VSS integration.
HPE Data Protector (9.08)
Page 9 of 119
Chapter 2: Integration concepts
The Data Protector VSS integration uses the Microsoft Volume Shadow Copy Service and Virtual Disk
Service interfaces and the VSSBAR agent to perform backup and restore of application data.
Volume Shadow Copy Service
Volume Shadow Copy Service (VSS) is a software service available on Microsoft Windows operating
systems.
During backup, the service collaborates with a backup application (for example, Data Protector), applications
to be backed up (usually database applications), shadow copy providers, and the operating system kernel to
create a consistent, point-in-time shadow copy set. These shadow copy sets can be backed up to backup
media or can be kept on a disk array for instant recovery or data mining.
During restore, the service collaborates with the backup application and database application to prepare for
the restore operation and to perform the recovery of the restored data.
The Data Protector Volume Shadow Copy integration provides a unified communication interface that can
coordinate backup and restore of an application regardless of their specific features. With this approach, a
backup application does not need to handle each application to be backed up specifically. However, the
production application as well as the backup application must conform to the VSS specification.
"Actors of the traditional backup model" on the next page and "Actors of the Data Protector VSS integration
backup model" on the next page show the differences between the traditional backup model and the model
with the Data Protector Microsoft Volume Shadow Copy Service integration.
HPE Data Protector (9.08)
Page 10 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 2: Integration concepts
Actors of the traditional backup model
Actors of the Data Protector VSS integration backup model
HPE Data Protector (9.08)
Page 11 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 2: Integration concepts
Without using the Volume Shadow Copy Service, Data Protector has to communicate with each
application to be backed up individually. The Data Protector VSS integration introduces a unified
backup and restore interface and provides the coordination among the participants of the backup and
restore process. The main actors of a VSS based backup are the:
l
Requestor – the backup application which requests a shadow copy creation.
l
Providers , which create the shadow copies. Two types of VSS providers exist:
l
l
software providers
l
hardware providers, supplied by the hardware (disk array) vendors
Writers – application-specific software modules which ensure application's data consistency when
a shadow copy is created.
Virtual Disk Service
Virtual Disk Service (VDS) is a Microsoft Windows service that provides a common interface for
managing storage hardware and disks and for creating volumes on these disks. Similar to VSS, the
hardware capabilities are abstracted by VDS providers.
Two types of VDS providers exist:
l
built-in software providers (at the operation system level)
l
hardware providers, supplied by the hardware (disk array) vendors.
You can use the VDS provider to configure and manage disks, volumes, partitions, and hardware RAID
subsystems. The disks can either be physical disks or virtual disks (LUNs). If your hardware providers
support this, you can manage individual physical disks in a LUN. The software provider does not
resolve LUNs to individual physical disks.
For some disk arrays, the Data Protector VSS integration can use the VDS hardware providers (if
present) to gather additional information needed for instant recovery and to perform the instant
recovery.
VSSBAR agent
The central part of the integration is the VSSBAR agent, which links Data Protector with the Microsoft
Volume Shadow Copy Service. Data Protector Microsoft Volume Shadow Copy Service integration
uses the VSSBAR agent for automatic browsing of VSS-aware writers, coordinating backup and
restore. The VSSBAR agent is responsible for the following actions:
l
detecting VSS writers
l
examining and analyzing Writer Metadata Document (WMD)
Writer Metadata Document (WMD) is metadata provided by each writer. Writers identify
themselves by the metadata and instruct the backup application what to back up and how to restore
the data. Thus, Data Protector follows the requirements provided by the writer when selecting the
volumes to be backed up and the restore method.
l
requesting shadow copy creation
HPE Data Protector (9.08)
Page 12 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 2: Integration concepts
l
backing up writers’ data to media
l
coordinating restore session start
l
restoring the Writer Metadata Document
l
restoring writer’s data from media
l
requesting instant recovery
During the Data Protector VSS integration backup, Data Protector does not interact directly with each
writer, but through the VSS interface. It uses the VSSBAR agent to coordinate the backup process.
The consistency of data is a responsibility of the VSS writer and not dependent on Data Protector
functionality. The shadow copy is created by the VSS provider.
Backup types
The available backup types depend on the type of the VSS provider used and its functionality:
l
l
With the software provider, you can back up the data to tape (standard Data Protector backup to
tape). This can be done only on the same system where the shadow copy is created.
With the hardware provider, you can create target volumes, which can be used for backing up to tape
or for instant recovery. Such a backup is a Data Protector zero downtime backup (ZDB), where
the hardware providers replace the disk array agent, which usually replicates the volumes. Thus, to
create a ZDB backup specification when working with the Data Protector VSS integration, you must
always select the hardware providers. For a general description of ZDB (split mirror or snapshot
backup) and instant recovery concepts, see the HPE Data Protector Concepts Guide.
The backup system to which the replicas are presented and from where they are backed up to tape
can either be:
l
A dedicated system (Data Protector ZDB terminology: a dual-host configuration).
In VSS terminology, a backup performed in such an environment is a transportable backup. It is
available only in combination with a hardware provider.
l
The same system as the application system (Data Protector ZDB terminology: a single-host
configuration).
In VSS terminology, a backup performed in such an environment is a local or network backup.
"Local VSS backup" on the next page shows the relationships between the components of a local
VSS backup. "Transportable VSS backup" on the next page shows the relationships between the
components of a transportable VSS backup.
HPE Data Protector (9.08)
Page 13 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 2: Integration concepts
Local VSS backup
Transportable VSS backup
VSS database
The VSS database (VSSDB) is an extension to the Data Protector Internal Database (IDB) on the Cell
Manager. It holds the VSS integration-specific information about the VSS backup sessions and their
HPE Data Protector (9.08)
Page 14 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 2: Integration concepts
replicas. The stored information can be used for the following purposes:
l
Instant recovery
l
Data mining (you can save backup components and writer metadata documents)
l
Listing the backup sessions and viewing details about them
l
Removing the backup sessions
l
Enabling (presenting and mounting) and disabling (dismounting and unpresenting) the ZDB-to-disk
and ZDB-to-disk+tape sessions' replicas
The VSSDB holds the VSS sessions' metadata. This metadata is stored in two parts of the VSSDB,
persistent part and the non-persistent part. The persistent part holds information about all backup
sessions, while the non-persistent part holds information only on instant recovery–enabled sessions
(ZDB-to-disk and ZDB-to-disk+tape sessions). Once a replica from the instant recovery-enabled
session is rotated out of the replica set, the information about the session is deleted from the nonpersistent part of the VSSDB and the session's target volumes are removed from the disk array. The
information about the session is still available in the persistent part of the VSSDB and can be deleted
only manually using the omnidbvss command. Note that ZDB-to-tape sessions and sessions created
using the software provider are always recorded only to the persistent part of the VSSDB.
You can query and manage the items of the VSSDB using the omnidbvss command. For information on
the command syntax, description, and examples, see the HPE Data Protector Command Line
Interface Reference.
The following tasks can be performed using the omnidbvss command:
l
"Querying the VSSDB" below
l
"Deleting backup sessions" below
l
"Enabling and disabling replicas" on the next page
Querying the VSSDB
Using the omnidbvss command, you can list:
l
All backup sessions, as well as their details.
l
All backup sessions based on a specific backup specification, as well as their details.
l
l
All backup sessions recorded in the persistent part of the VSSDB, which were created before a
specified date, as well as their details.
Details on a specific backup session, identified by the session ID.
Using the omnidbvss command, you can save the documents about the backup components and writer
metadata to a specified directory.
See the omnidbvss man page for the command syntax and examples.
Deleting backup sessions
Using the omnidbvss command, you can delete:
l
A specific instant recovery-enabled backup session (a replica version), identified by the session ID,
from the non-persistent part of the VSSDB and disk array, or only from the VSSDB.
HPE Data Protector (9.08)
Page 15 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 2: Integration concepts
l
l
l
All instant recovery-enabled backup sessions based on a specific backup specification (a replica
set) from the non-persistent part of the VSSDB and disk array, or only from the VSSDB.
A specific backup session, identified by the session ID, from the persistent part of the VSSDB.
All backup sessions based on a specific backup specification or session that were created before a
specified date from the persistent part of the VSSDB.
See the omnidbvss man page for the command syntax and examples.
Enabling and disabling replicas
Using the omnidbvss command, you can present and mount (enable) on a backup system and
dismount and unpresent (disable) from a backup system the replicas from:
l
All instant recovery-enabled sessions.
l
A specific instant recovery-enabled session, identified by the session ID.
l
All instant recovery-enabled sessions based on a specific backup specification.
Changes in the environment
During VSS instant recovery sessions or VSSDB maintenance sessions using the omnidbvss
command (remove, disable, enable), the state of target volumes in a replica is checked prior to restore,
remove, disable, and enable operations. Before restore, the state of the source volumes on the
application system is also checked. If the check finds any changes that have not been tracked in the
VSSDB, the session is aborted, notifying you about the specific changes. This check can be disabled
by setting the OB2VSS_IGNORE_BACKUP_DISK_CHANGES and OB2VSS_IGNORE_SOURCE_DISK_CHANGES
omnirc options to 1.
During a VSS backup session, a replica is created and left on a disk array until the specified number of
replicas rotated is reached. After that, the next replica to be created replaces the oldest replica in the
set. If the oldest replica in the set cannot be replaced because you have manually (not using Data
Protector) unpresented the replica from the backup system and presented it on some other system, the
backup session continues and creates a new replica and leaves the old one on the array. Note that in
this case, if there is no space on the disk array for creating a new replica, the backup session also fails.
If you set the OB2VSS_IGNORE_BACKUP_DISK_CHANGES omnirc option to 1, the backup session
continues by removing the old replica and creating a new one in its place.
HPE Data Protector (9.08)
Page 16 of 119
Chapter 3: Configuration
Prerequisites and limitations
This is a list of prerequisites and limitations for the Data Protector Microsoft Volume Shadow Copy Service
integration. Additional limitations and recommendations that are not directly connected to the integration
(such as operating system and GUI limitations) and disk array limitations are listed in the HPE Data Protector
Product Announcements, Software Notes, and References.
Prerequisites
l
Before you begin, ensure that you have correctly installed and configured Data Protector, writers. If you
want to perform ZDB, install and configure also the VSS and VDS hardware providers. See the:
l
Latest support matrices on the web for an up-to-date list of supported versions, platforms, devices, disk
arrays, limitations, and other information.
https://softwaresupport.hpe.com/
l
l
l
HPE Data Protector Installation Guide for information of how to install Data Protector on various
architectures and how to install the Data Protector Microsoft Volume Shadow Copy Service integration
and ZDB agents.
Writers and shadow copy providers documentation for instructions of how to install and configure
writers and providers on your system.
Install any needed Microsoft Windows patches or hotfixes. For a detailed list, see the latest support
matrices at https://softwaresupport.hpe.com/.
Transportable backup prerequisites
l
l
The backup system must be configured to accept connections from the application system and the other
way round.
The following Data Protector and HPE storage components must be installed and configured on both the
application and backup system:
l
MS Volume Shadow Copy Integration
l
The appropriate Data Protector disk array agent
l
VSS hardware provider
Depending on the operating system version and disk array, the following HPE Storage components may
be required:
l
VDS hardware provider
o
For ZDB, the VDS hardware provider is optional.
o
For instant recovery, certain limitations may apply, depending on the disk array family and version of
HPE Data Protector (9.08)
Page 17 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
the operating system. In some cases, the VDS hardware provider is required for a successful
instant recovery. See "Instant recovery" on page 48.
Limitations
For a list of general Data Protector limitations, see the HPE Data Protector Product Announcements,
Software Notes, and References.
l
l
l
To run a VSS integration backup, the writer’s data must be on an NTFS filesystem. For hardware
providers, this is not required.
The VSS integration backup of writers which store their data on network shared volumes is not
supported.
The Data Protector Microsoft VSS integration does not provide any restore method for writers
requesting a custom restore method. These writers are by default not presented by Data Protector.
If a writer specifies a custom restore method, it is only possible to restore the writer's data as plain
files using the Data Protector functionality. You can perform the custom restore manually. For
additional information on the restore methods, see the writer's documentation.
l
Preview is not possible for VSS backup and restore sessions.
l
Backup preview is only available for VSS filesystem backup sessions.
l
Dynamic disks (Logical Disk Manager partitions) can only be backed up with software providers.
ZDB and instant recovery limitations
l
l
l
l
Not all VSS writers supported by Data Protector can be used with the instant recovery functionality.
For a list of the VSS writers that support the instant recovery functionality, see the latest support
matrices at https://softwaresupport.hpe.com/.
VSS backups with hardware provider are only supported for volumes on disk arrays. This is also
valid for system disks. Instant recovery of system disks backups is not possible.
The application system and the backup system that you use must have the same operating system
version installed.
During a local backup, with the option Track the replica for instant recovery selected, and
with the option Mount the replica not selected, note the following:
l
l
On Windows Server 2003 systems, the created snapshots are unpresented on the application
system after the backup.
On Windows Server 2008 systems, the created snapshots are only put offline after the backup,
while they remain presented on the application system.
HPE P6000 EVA Disk Array Family hardware provider limitations
l
On Windows Server 2003 systems if a disk array from the HPE P6000 EVA Disk Array Family is
used, instant recovery is possible only if the VDS hardware providers is installed.
On Windows Server 2008 systems, the VDS providers are optional. However, if the VDS provider
was installed the time of the backup session, it must be installed at the time of the instant recovery
session.
HPE Data Protector (9.08)
Page 18 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
HPE P9000 XP Disk Array Family hardware provider limitations
l
If a disk array from the HPE P9000 XP Disk Array Family is used and the P9000 XP Array VSS
provide is in resync mode, the VDS providers are optional for instant recovery. However, if the VDS
provider was installed the time of the backup session, it must be installed at the time of the instant
recovery session.
If the P9000 XP Array provider is in VSS compliant mode, the VDS providers are always required for
instant recovery.
l
l
l
You cannot switch between the VSS compliant mode and resync mode during execution of a backup
session.
If the P9000 XP Array provider is in resync mode, a maximum of three replicas (S-VOLs) can be
created for a source volume (P-VOL).
If the HPE P9000 XP Disk Array Family VSS hardware provider is in the VSS compliant mode, and
there is a high read-write load on the disk array, the replica creation may take a long time.
Consequently, the backup session may fail.
By default, Data Protector waits for a maximum of 45 minutes (2700 seconds) for mirrors to become
synchronized. If synchronization is not completed when the waiting period expires, Data Protector
aborts the backup session.
You can increase the waiting period duration by setting the OB2VSS_WAIT_TIMEOUT omnirc option.
The option value determines the period (in seconds) for which the Data Protector agent waits for
mirrors that are created in one session to become fully synchronized. If the OB2VSS_WAIT_TIMEOUT
option is set to more than 2 hours (7200 seconds), you should additionally increase the value of the
global option SmIdleTimeout to be greater that the value of OB2VSS_WAIT_TIMEOUT. The value of
SmIdleTimeout should be specified in minutes, rather than seconds.
If backup sessions are lengthy and there is a possibility that the VSS backup window will not be
met, it is highly recommended that you first switch the HPE P9000 XP Disk Array Family VSS
hardware provider to resync mode. Afterwards, consecutively run backup sessions until the backup
session count reaches the current value of the Number of replicas rotated option. This will start
creating disk mirrors on P9000 XP Array that will be used in scheduled backups.
l
Second-level mirrors are not supported.
For supported backup and connectivity topologies, see the HPE Data Protector Zero Downtime
Backup Administrator's Guide.
Configuring the application system for instant
recovery-enabled backup sessions
To be able to perform instant recovery-enabled backups and then instant recovery, you must resolve
the application system using the omnidbvss command. During the resolve operation, Data Protector
contacts VDS to get information about the storage IDs of the source volumes presented on the
application system. The information about the storage IDs is required during instant recovery, in case
such information cannot be gathered at restore time.
If resolving is not performed before a backup session, one of the following happens:
HPE Data Protector (9.08)
Page 19 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
l
l
l
The application system is resolved automatically during the backup session, if the OB2VSS_
DISABLE_AUTO_RESOLVE option in the omnirc file is set to 0 (default). In this case, the backup time
for creating a replica is prolonged.
A ZDB-to-disk session fails notifying you to resolve the application system before the backup is run,
if the OB2VSS_DISABLE_AUTO_RESOLVE option is set to 1.
A ZDB-to-disk+tape session completes with the warning that only backup to tape was performed not
leaving the replica on the disk array and thus disabling instant recovery, if the OB2VSS_DISABLE_
AUTO_RESOLVE option is set to 1.
Always perform the resolve operation after:
l
l
l
Installing or upgrading Data Protector.
Your source volumes' configuration on the application system has changed (for example, you have
modify the existing source volumes or you have presented new source volumes).
You have added a new storage object (for example, a Microsoft Exchange Server storage group).
To resolve the application system, execute the following on any VSS client in the Data Protector cell:
omnidbvss –resolve {-apphost ApplicationSystem | –all}
where ApplicationSystem is the name of the application system you want to resolve. In a cluster
environment (for example, Microsoft Exchange Server CCR), provide the virtual server name for the
ApplicationSystem parameter to enable the resolve operation on both cluster nodes.
To resolve all application systems in a cell simultaneously, use the option –all.
Alternatively, you can set the omnirc option OB2VSS_ALWAYS_RESOLVE_SOURCES to 1. In this case, the
source volumes on the application system are resolved automatically during every instant recoveryenabled backup session. However, the replica creation time is considerably prolonged. Do not set this
option to 1 in a CCR Microsoft Exchange environment if you plan to back up only the database copy
disks and leave the possibility for eventual instant recovery from these disk to the production database
disks. In this case, the production database disks are never resolved during backup and this can cause
such an instant recovery to fail.
Configuring HPE P4000 SAN Solutions
This section describes how to configure the Data Protector HPE P4000 SAN Solutions integration for
use with the VSS integration.
Prerequisites
l
Obtain or install HPE storage licenses and components:
l
HPE P4000 SAN Solutions VSS provider.
l
HPE P4000 SAN Solutions DSM (Device Specific Module) for MPIO.
For details of how to install these components, see the HPE P4000 Windows Solution Pack
documentation.
HPE Data Protector (9.08)
Page 20 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
For information on supported product versions, see the latest support matrices at
https://softwaresupport.hpe.com/.
l
Obtain or install Data Protector licenses and components:
l
An instant recovery license.
l
HPE P4000 VSS Agent.
For installation and licensing information, see the HPE Data Protector Installation Guide.
l
l
l
Make sure the same operating system (and its version) is installed on the application and backup
systems.
Connect an HPE P4000 SAN Solutions storage system to the application and backup systems
through the SAN. Backup system must be connected to the same SAN as the HPE P4000 SAN
Solutions storage system.
For VSS transportable backups, the source volumes must be presented to both the application and
backup systems.
For supported backup and connectivity topologies, see the HPE Data Protector Zero Downtime
Backup Administrator's Guide.
For general Data Protector and integration-specific limitations, see the HPE Data Protector Product
Announcements, Software Notes, and References.
Configuring the integration
Before you start configuration, make sure you met the prerequisites described in "Prerequisites" on the
previous page. Then set the login information for the SMI-S HPE P4000 SAN Solutions provider
running on a management system.
To set, delete, list, or check the login information, use the omnidbp4000 -–ompasswd command. For
command syntax and examples, see the HPE P4000 SAN Solutions part of the HPE Data Protector
Zero Downtime Backup Administrator's Guide and the omnidp4000 reference page in the HPE Data
Protector Command Line Interface Reference.
The information you provide is kept in the HPE P4000 SAN Solutions part of the ZDB database for
SMI-S based integrations (SMISDB).
For each management system you configure, the following information is stored:
l
l
Hostname as recognized in the IP network.
Port number through which HPE P4000 VSS Agent communicates (default - 5988). If the SMI-S
HPE P4000 SAN Solutions provider accepts the SSL-based connection, the default port number is
5989.
NOTE: When using StoreVirtual VSA software version 12.5, use SSL-based connection.
l
Management group name, and management group username and encrypted password of an HPE
P4000 SAN Solutions provider login account.
The HPE P4000 SAN Solutions part of the SMISDB resides on the Cell Manager in the directory Data_
Protector_program_data\server\db80\smisdb\p4000 (Windows systems) or
/var/opt/omni/server/db80/smisdb/p4000/login (UNIX systems).
HPE Data Protector (9.08)
Page 21 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
It is recommended to use the default port number (5988 for the non-SSL and 5989 for the SSL-based
SMI-S P4000 SAN Solutions provider connection settings).
It is also recommended to execute omnidbp4000 --ompasswd --check [--host ClientName]
before backup or instant recovery to verify the configuration of SMI-S HPE P4000 SAN Solutions
provider.
Configuring the VSS hardware provider
Set the management group credentials, on both the application and the backup system. The credentials
required for using the HPE P4000 SAN Solutions VSS provider are:
l
Management group name
l
Management group user name and password
Depending on the operating system, use either the Authentication Console or edit the Windows
registry. See the HPE P4000 SAN Solutions VSS provider documentation for details.
After you complete this procedure, the HPE P4000 SAN Solutions VSS provider is ready to be used for
zero downtime backup and instant recovery.
Configuring HPE P6000 EVA Disk Array Family
This section describes how to configure the Data ProtectorHPE P6000 EVA Disk Array Family (P6000
EVA Array) integration for use with the VSS integration.
Prerequisites
l
Obtain or install HPE storage licenses and components:
l
HPE Virtual Controller Software (VCS or XCS) and Command View (CV) EVA.
For installation instructions, see the VCS and CV EVA documentation. For information on
supported product versions, see the latest support matrices at https://softwaresupport.hpe.com/.
l
l
HPE Business Copy (BC) P6000 EVA microcode and license.
l
HPE MPIO Full Featured DSM (Device Specific Module) for HPE P6000 EVA Disk Array Family.
l
A P6000 EVA Array license for, at least, basic and snapshot operation.
Obtain or install Data Protector licenses and components:
l
An instant recovery license.
l
HPE P6000 / HPE 3PAR SMI-S Agent.
If this agent is not installed, only “Restore using Microsoft Virtual Disk Service” (switch of disks)
is possible. For installation and licensing information, see the HPE Data Protector Installation
Guide.
HPE Data Protector (9.08)
Page 22 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
l
l
Make sure the same operating system (and its version) is installed on the application and backup
systems.
Connect P6000 EVA Array to the application and backup systems through the SAN. Backup system
must be connected to the same SAN as the P6000 EVA Array.
For supported backup and connectivity topologies, see the HPE Data Protector Zero Downtime
Backup Administrator's Guide.
For general Data Protector and integration-specific limitations, see the HPE Data Protector Product
Announcements, Software Notes, and References.
Configuring the integration
Before you start configuration, make sure you met the prerequisites described in "Prerequisites" on the
previous page. Then set the login information for SMI-S P6000 EVA Array provider running on a
management system.
To set, delete, list, or check the login information, use the omnidbsmis -ompasswd CLI command. For
command syntax and examples, see the HPE Data Protector Command Line Interface Reference.
The information you provide is kept in the ZDB database for the HPE P6000 EVA Disk Array Family
integration (SMISDB).
For each management system you configure, the following information is stored:
l
l
l
Hostname as recognized in the IP network.
Port number through which P6000 EVA SMI-S Agent communicates (default - 5988). If the SMI-S
P6000 EVA Array provider accepts the SSL-based connection, the default port number is 5989.
User name and encoded password for the SMI-S P6000 EVA Array provider login.
SMISDB resides on the Cell Manager in the directory Data_Protector_program_
data\server\db80\smisdb (Windows systems) or /var/opt/omni/server/db80/smisdb (UNIX
systems).
Considerations
If a failover from the active to the standby management system happens, proceed as follows:
l
l
If standby and failed management systems have the same hostname, no action is needed.
If standby and failed management systems have different hostnames, remove the failed system
from the Data Protector configuration, and then add the new management system.
It is recommended to use the default port number (5988 for non-SSL and 5989 for the SSL-based SMI-S
P6000 EVA Array provider connection settings).
It is also recommended to execute omnidbsmis -ompasswd -check [-host ClientName] before
backup or instant recovery to verify the configuration of SMI-S P6000 EVA Array provider.
Configuring the VSS hardware provider
Open the HPE EVA VSS Configuration Utility and follow the steps below:
HPE Data Protector (9.08)
Page 23 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
1. Ensure that the correct management appliance IP address is specified.
2. Click Login and in the Login dialog box, enter the username and password.
3. Select SnapClone as the Snapshot Type.
4. Click Select Disk Group and select the disk group with the disks that you intend to back up for
instant recovery.
Open the HPE EVA VDS Configuration Utility and follow the steps below:
1. Ensure that the correct Appliance IP address is specified.
2. Click Login and in the Login dialog box, enter the username and password.
3. Click Select Disk Group and select the following disk groups:
l
The disk group with disks created by the VSS provider.
l
The disk group with the source disks.
After you complete this procedure, the VSS and VDS providers are ready to be used for zero downtime
backup and instant recovery.
Configuring HPE P9000 XP Disk Array Family
This section describes how to configure the Data ProtectorHPE P9000 XP Disk Array Family
integration for use with the Data Protector VSS integration.
Prerequisites
l
Obtain or install HPE storage licenses and components:
l
RAID Manager Library on the application and backup systems. For installation instructions, see
the RAID Manager Library documentation.
RAID Manager Library is firmware-dependent. Consult the HPE sales representative for
information on which version of RAID Manager Library to use.
l
HPE Business Copy (BC) P9000 XP microcode and a license for at least basic and HPE
Business Copy operation.
l
P9000 XP Array VSS and VDS provider.
l
HPE MPIO Full Featured DSM (Device Specific Module) for HPE P9000 XP Disk Array Family
For information on supported product versions, see the latest support matrices at
https://softwaresupport.hpe.com/.
l
Obtain or install Data Protector licenses and components:
l
An instant recovery license.
l
The HPE P9000 XP Agent.
HPE Data Protector (9.08)
Page 24 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
If this agent is not installed, it is not possible to restore data from the P9000 XP Array. For
installation and licensing information, see the HPE Data Protector Installation Guide.
l
Make sure the same operating system and version is installed on both application and backup
systems.
l
Connect the application and backup systems to the same P9000 XP Array.
l
Assign LUNs to the respective ports.
l
Pre-configure the LDEVs for the VSS snapshot creation and put them in the S-VOL host group. If
you perform restore using VDS, you need to put new LDEVs in the S-VOL host group after the
restore, since VDS restore switches the disks and the replica becomes the source volume. Add as
many new LDEVs as were used for restore.
Configuring the VSS hardware provider
Before running ZDB sessions, open the HPE XP VSS Hardware Provider Configuration Utility:
l
Choose the configuration mode for your backups: VSS Compliant Mode or Resync Mode.
Note: Changing the modes between backup sessions may have an impact on restore. For
details, see "HPE P9000 XP Array considerations" on page 54.
l
Enable the use of snapshots (if supported by the P9000 XP Array and the VSS provider), select the
Snap Type snapshot.
Data Protector allows both, backups in resync mode and with snapshots for the same source
volumes.
For details, see the HPE P9000 XP Disk Array Family VSS hardware provider documentation.
Configuring the user authentication data
Before running zero downtime backup (ZDB) and instant recovery (IR) sessions on a disk array which
is operating in the user authentication mode, add appropriate user credentials to the ZDB database
(XPDB). For more information on the user authentication mode and on how to configure disk array user
authentication data, see the HPE Data Protector Zero Downtime Backup Administrator's Guide.
Configuring HPE 3PAR StoreServ Storage
This section describes how to configure the Data ProtectorHPE 3PAR StoreServ Storage integration
for use with the Data Protector Microsoft Volume Shadow Copy Service integration. Before you start
with the configuration, ensure the prerequisites discussed in the following section are fulfilled.
Prerequisites
l
Obtain or install HPE storage licenses and components:
l
HPE 3PAR VSS Provider Software.
HPE Data Protector (9.08)
Page 25 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 3: Configuration
For installation instructions, see the HPE 3PAR StoreServ Storage documentation. For information
on supported product versions, see the latest support matrices at https://softwaresupport.hpe.com/.
l
Obtain or install Data Protector licenses and components:
l
l
Appropriate zero downtime backup extension and instant recovery extension licenses-to-use
(LTU).
HPE 3PAR VSS Agent installed on both the application system and the backup system
For licensing and installation information, see the HPE Data Protector Installation Guide.
l
l
l
Make sure the same operating system (and its version) is installed on the application and backup
systems.
Connect a storage system of the HPE 3PAR StoreServ Storage family to the application and
backup systems through the SAN. The backup system must be connected to the same SAN as the
storage system of the HPE 3PAR StoreServ Storage family.
Source volumes must have snapshot space (copy space) in a storage system's Common
Provisioning Group (CPG) associated with.
For supported backup and connectivity topologies, see the HPE Data Protector Zero Downtime
Backup Administrator's Guide.
For general Data Protector and integration-specific limitations, see the HPE Data Protector Product
Announcements, Software Notes, and References and the HPE Data Protector Zero Downtime
Backup Administrator's Guide.
Configuring the Data Protector HPE 3PAR StoreServ
Storage integration
For configuration instructions, see the HPE Data Protector Zero Downtime Backup Administrator's
Guide, part HPE 3PAR StoreServ Storage, chapter Configuration.
Configuring the HPE 3PAR VSS hardware provider
The only step you need to perform is to configure a VSS hardware provider user account on both the
application system and the backup system.
The provider user account must have the same privileges as the user account that you added while
configuring the Data ProtectorHPE 3PAR StoreServ Storage integration: the Edit privilege level on the
application system and the source volumes. Such a user account ensures the VSS hardware provider
proper access to the 3PAR StoreServ Storage. The configured user accounts on the application and
backup systems can be different. For instructions, see the HPE 3PAR StoreServ Storage
documentation.
After you complete this process, the HPE 3PAR StoreServ Storage VSS provider is ready to be used in
Data Protector zero downtime backup and instant recovery sessions.
HPE Data Protector (9.08)
Page 26 of 119
Chapter 4: Backup
Backup types
VSS backup types
The following VSS backup types are available with the Data Protector VSS integration:
l
Local backup
l
Transportable backup
See "Backup types" on page 13 for more details on VSS backup types.
Application specific backup types
For supported application specific backup types, such as Full, Differential, and so on, see the
appropriate writer specific sections in "Writer specifics" on page 62.
Backup flow
1. The coordinator (the VSS service) identifies all writers that support the VSS feature and passes the list
of available writers and their characteristics (Writer Metadata Document) back to Data Protector.
2. Data Protector examines Writer Metadata, identifies the volumes that contain the data to be backed up,
and prepares a list of volumes (shadow copy set) that must be put into a consistent state and passes
this information back to the coordinator, which informs available writers.
3. The VSSBAR agent notifies the writers about the shadow copy creation. The VSS mechanism ensures
that there are no writes on the volume while the shadow copy is being created.
The VSSBAR agent then passes the shadow copy creation requests to VSS.
4. After a shadow copy is created, the VSS service returns the related information to Data Protector. If an
instant recovery enabled ZDB is performed, the VSSBAR agent coordinates the VDS agent to gather
information on the created replica that is needed for instant recovery.
5. Data Protector backs up the data from the shadow copy to media and then notifies the VSS service that
the shadow copy can be released. The shadow copy provider destroys the shadow copy that has been
already backed up.
"Local VSS backup" on page 14 shows the relations between the actors of a local VSS backup.
6. Zero downtime backup:
The backup scenarios, which are not mutually exclusive, depend on the:
l
backup options selected in the backup specification
l
the ZDB type selected in the backup specification or at the start of the backup
l
the VSS backup type (local or transportable).
The replica can be:
HPE Data Protector (9.08)
Page 27 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
l
Kept on a disk array for rotation and instant recovery purposes.
l
Kept on a disk array only for data mining.
l
l
l
l
Moved to a backup medium from the application system (local backup) or backup system
(transportable backup).
Presented and mounted to the backup system specified in the backup specification. It can be
mounted in read–only or read/write mode.
If not kept on the array, destroyed after the backup session completes (after data in the replica
is moved to a backup medium).
Manipulated using the Data Protector omnidbvss command.
7. The related information about the completed backup session is recorded in the VSSDB.
Zero downtime backup with the Data Protector
VSS integration
The following ZDB types are available with the Data Protector VSS integration:
l
ZDB to tape
l
ZDB to disk
l
ZDB to disk+tape
For more details on ZDB types, see the HPE Data Protector Concepts Guide.
HPE P4000 SAN Solutions
There is one replica type available when backing up using the HPE P4000 SAN Solutions hardware
providers:
l
Snapshot
The snapshot is dependent on previous snapshots and the original volume—when a snapshot is
created, instead of writing to an original volume, new writes are redirected to a new location (“redirect
on write”). When a resync (instant recovery) of a snapshot is performed, all dependent snapshots
are automatically deleted.
HPE P6000 EVA Disk Array Family
There are three replica types available when backing up using the P6000 EVA Array hardware
providers:
l
l
Snapshot with pre-allocation of disk space (standard snapshot). Snapshots can be created from
original volumes or mirrorclones.
Snapshot without pre-allocation of disk space (vsnap or virtually capacity-free snapshot). Vsnaps
HPE Data Protector (9.08)
Page 28 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
can be created from original volumes or mirrorclones.
l
A full copy of the source volume (original virtual disk), independent of the original virtual disk
(snapclone).
In the Data Protector GUI, you can select between snapshots or vsnaps (Differential (Snapshot)) and
snapclones (Plex (Clone/Mirror)). To create vsnaps instead of standard snapshots, you must first
configure the hardware provider to create vsnaps.
If you select Keep the replica after the backup or enable instant recovery (by selecting the Track the
replica for instant recovery), you can further select the subtypes:
l
l
Differential (Snapshot):
l
Snapshot
l
MirrorClone snapshot
Plex (Clone/Mirror):
l
Snapclone
HPE P9000 XP Disk Array Family
There are two replica types available:
l
Mirrors (Plex (Clone/Mirror))
l
Snapshots (Differential (Snapshots))
Snapshots must be supported by the P9000 XP Array and the P9000 XP Array VSS hardware
provider.
Additionally, there are two configuration modes available for the backup with the P9000 XP Array
hardware providers:
l
VSS compliant mode
When the P9000 XP Array provider is in VSS compliant mode, the source volume (P-VOL) and its
replica (S-VOL) are in simplex, unpaired state after backup. Therefore the number of replicas (SVOLs per P-VOL) rotated is not limited. Restore from a backup in such a configuration is possible
only by switching disks.
Note that if you select Differential (Snapshots) as the replica type, the VSS compliant mode is
ignored and the resync mode is always used.
l
Resync mode
When the P9000 XP Array provider is in the resync mode, the source volume (P-VOL) and its replica
(S-VOL) are in a suspended mirror relationship after backup. The maximum number of replicas (SVOLs per P-VOL) rotated is three, provided that the MU range is 0-2 or 0, 1, 2. Restore from a
backup in such a configuration is possible only by re-synchronization of an S-VOL with its P-VOL.
HPE 3PAR StoreServ Storage
There is one replica type available when backing up using the HPE 3PAR StoreServ Storage hardware
providers:
HPE Data Protector (9.08)
Page 29 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
l
Snapshot
The snapshot is independent of previous snapshots—after it is created, new writes are not
redirected to a new location but continue to be made on the original volume after its current contents
are copied over (“copy on write”).
Considerations
Microsoft Cluster environments and transportable
backups
When using a VSS hardware provider in a Microsoft Cluster environment, a transportable shadow copy
(replica) must be transported outside of the cluster if the original volume is mounted within the cluster.
As a result, Data Protector supports only the following configurations in a cluster:
l
Local or network backup using a VSS software provider.
l
Transportable backup, where the backup system is not part of the cluster (not a cluster node).
HPE P4000 SAN Solutions
When a new snapshot is created with HPE P4000 SAN Solutions, new disk space is allocated on the
array where the new data is written (this method of creating a snapshot is known also as redirect on
write). This means that newer snapshots depend on older ones and that you need all snapshots in a
chain for the last snapshot to be valid.
If several backup specifications contain the same source volume, instant recovery from a ZDB session
of one backup specification may cause that the newer ZDB sessions based on other backup
specifications are not available for instant recovery.
For example, the backup specification BSpec1 contains volumes C: and G: and the backup
specification BSpec2 volumes G: and M:.
An instant recovery from a ZDB session of BSpec1 will remove all subsequent snapshots of the
volumes C: and G:. This will also prevent further instant recovery sessions from all ZDB sessions of
BSpec2 that were created after the ZDB session for BSpec1 that is being used for instant recovery,
since the snapshots for the volume G: are no longer available.
As a result, you must consider the following:
l
l
l
It is recommended to limit the number of volumes that are included in the backup specification to a
minimum needed.
Perform instant recovery for sessions and not for particular volumes.
Any newer replica (snapshot or smartclone) created outside of Data Protector, will prevent instant
recovery from a session that contains the volume which was replicated outside of Data Protector.
HPE Data Protector (9.08)
Page 30 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
HPE P6000 EVA Disk Array Family
l
l
When the cloning process for a source volume is in progress, another replica of the same source
volume cannot be created.
Only one type of target volumes per source volume can exist on a P6000 EVA Array at the same
time. For example, creation of a snapclone may fail if a snapshot or vsnap for the same source
volume exists on the array. You should first delete such replicas using the Data Protector command
omnidbvss, or using Command View EVA if they were not created by Data Protector.
Note that in case of a MirrorClone Snapshot, the omnidbvss command can delete only the
snapshots. The MirrorClone used as the source for the snapshot cannot be deleted using omnidbvss
and must be deleted manually using EVA Command View, by first detaching the MirrorClone and
then deleting the VDisk.
l
l
On Windows Server 2003 systems, the VDS hardware provider is required for instant recovery.
On Windows Server 2003 systems, when the VDS hardware provider is not used, the backup
system is left enabled (in read/write mode) even if the option Enable backup system in read/write
mode was not selected. As a result, you must ensure that you do not delete or break these shadow
copies when working with other (third-party) tools.
Replica creation and reuse
A new replica is created and added to the replica set when the specified Number of replicas rotated
is not reached.
The oldest replica in the set is deleted and the new one is created when the specified Number of
replicas rotated is reached.
HPE P9000 XP Disk Array Family
l
For backups (mirrors) created in the resync mode, the value of the Number of replicas rotated
option is limited to a maximum of three replicas (S-VOLs) per source volume (P-VOL), provided that
the MU# range option in the P9000 XP Array VSS hardware provider configuration is set to 0-2 or 0,
1, 2. When the same P-VOL is used in more than one backup specification, the sum of values
specified in the Number of replicas rotated option in all backup specifications should not exceed
3.
This limitation also applies for the following backup scenario in a Microsoft Exchange Server CCR
environment with the same backup system selected: Suppose you create three backups of a source
volume on the production database system with one backup specification, and then create three
backups of the source volume on the database copy system with a different specification. Together,
this would create six replicas. However, Data Protector limits the number and supports only three
replicas together in such a case. Therefore, you can choose to back up source volumes only on the
production database system or only on the database copy system. Otherwise, after three replicas
are created with one backup specification, the next backup with the second backup specification will
fail.
Snapshots are not limited to 3 replicas per source volume. The MU# number for snapshots can be
configured in the VSS provider configuration.
HPE Data Protector (9.08)
Page 31 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
l
You can change the P9000 XP Array VSS hardware provider mode between different backup
sessions using the same backup specification, but this is not recommended when you use replica
set rotation, because restore of such backup data may fail.
For details, see "HPE P9000 XP Array considerations" on page 54.
l
On Windows Server 2003 systems, when the VDS hardware provider is not used, the backup
system is left enabled (in read/write mode) even if the option Enable backup system in read/write
mode was not selected. As a result, you must ensure that you do not delete or break these shadow
copies when working with other (third-party) tools.
HPE 3PAR StoreServ Storage
l
l
After a new snapshot is created with HPE 3PAR StoreServ Storage, the data on the original volume
which is about to be overwritten by new data is first copied to the snapshot space associated with
the volume (this method of creating a snapshot is known also as copy on write). This means that
newer snapshots are independent of older ones.
HPE recommends not to manually set retention periods on target volumes created by Data
Protector. Volume retention period, when set, prevents a volume from being deleted from the 3PAR
StoreServ Storage when Data Protector rotates it out of the replica set. In this case, after replica set
rotation, the volume becomes an orphaned volume that unnecessarily occupies storage space. Due
to a problem in the HPE 3PAR VSS hardware provider version 2.1.0.11, failed deletions of such
volumes are reported as successful.
Configuration check
With instant recovery, it is possible to selectively restore separate writer’s components if they reside
on separate source volumes. Instant recovery of separate components also requires that the volumes
with components data should not contain any other data. The configuration check detects, whether
there is more than one component on the volume and whether there is any other data besides the
component’s data.
At backup time, Data Protector can check whether the individual components can be selectively
restored using the instant recovery functionality. At instant recovery time, Data Protector can check
whether the data required for the components restore is available.
In case of the Microsoft Exchange Server Writer, it is checked, whether the whole storage group should
be restored, or whether the separate database stores can be restored individually. In case of the MSDE
Writer, it is checked, whether the user, system, and log files are on separate volumes.
If the check fails for a component during the backup, you will not be able to select this component for
the instant recovery, you will need to recover all writer components. If the check fails during the instant
recovery, the instant recovery session will fail.
Configuration check modes
You can select between three configuration check modes:
l
Strict
If any file or folder on the volume does not belong to the component, the ZDB to disk, ZDB to
disk+tape, or instant recovery fails.
HPE Data Protector (9.08)
Page 32 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
l
Non-strict
If any folder on the volume does not belong to the component, the ZDB to disk, ZDB to disk+tape, or
instant recovery fails.
l
Disabled
The check detects whether there is more than one component on the volume or there is any other
data besides the component’s data on the volume, but the session does not fail in any case.
If the configuration check is disabled for an instant recovery, you will lose the data that does not belong
to a component, but resides on the same volume as the component. Disable configuration check only if
instant recovery cannot be performed with an enabled configuration check and only after you make sure
that this will not result in a loss of data. In case of a data loss, the data that does not belong to a
component, but resides on the same volume, will be lost.
The configuration check should not be disabled except for the cases when instant recovery cannot be
performed with an enabled configuration check. Due to the specific behavior, some writers (not
applicable to the MSDE Writer and Microsoft Exchange Server writers) may create temporary files on
components volumes during backup and instant recovery causing the check failure. In such cases,
instant recovery is not possible without disabling the check.
Creating backup specifications
The procedure below shows how to back up Microsoft VSS objects using the Data Protector GUI.
Some writers have specific limitations. For writers specific prerequisites, limitations, and additional
steps see the appropriate sections in "Writer specifics" on page 62.
To create a new backup specification for the VSS integration, proceed as follows:
1. In the HPE Data Protector Manager, switch to the Backup context.
2. In the Scoping Pane, expand Backup Specifications.
3. Right-click MS Volume Shadow Copy Writers and click Add Backup.
4. In the Create New Backup dialog box, select the backup type. You can choose between the
following types:
l
Local or network backup
This type is used for single-host VSS backup. To perform a zero downtime backup (ZDB), you
need a hardware provider. Otherwise, no hardware provider is required for this type of backup.
l
VSS transportable backup
Use this option to create shadow copies on the application system and present them to the
backup system, which can perform the backup to tape.
A hardware provider is required for this type of backup.
HPE Data Protector (9.08)
Page 33 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
Selecting VSS transportable backup
5. In Application system, specify the name of the client that has the VSSBAR agent installed.
When backing up cluster-aware writers (such as SQL Server via the MSDE Writer, or Exchange
Server in LCR or CCR environment), specify the virtual server name given in the particular writer
resource group.
Zero downtime backup:
Specify the following options:
l
l
l
l
l
For local or network backup, select Use hardware provider to enable ZDB and specify other
backup options.
For a transportable backup, the name of the backup system from where the shadow copy can
be backed up to tape or where you want your shadow copies to be presented and mounted after
the backup. The hardware provider is used automatically.
To keep the replica after backup, select Keep the replica after backup.
To enable instant recovery, select Track the replica for instant recovery. Keep the replica
after backup is selected automatically.
If you select Keep the replica after the backup, you can specify the replica type:
Replica types
HPE Data Protector (9.08)
Page 34 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
Replica Type
Description
Plex
A shadow copy, independent from its source volume is created:
(Clone/Mirror) o This replica type is not supported by the HPE P4000 SAN Solutions
and HPE 3PAR StoreServ Storage hardware providers.
Differential
(Snapshot)
Default
o
With the HPE P6000 EVA Disk Array Family hardware provider, a
snapclone is created.
o
With the HPE P9000 XP Disk Array Family hardware provider, the
replica created depends on the provider mode (clone in the case of VSS
compliant mode, or mirror in the case of resync mode). The selected
subtype is ignored.
A shadow copy, dependent on its source volume is created.
o
With the HPE P4000 SAN Solutions or HPE 3PAR StoreServ Storage
hardware provider, a snapshot is created.
o
With the HPE P6000 EVA Disk Array Family hardware provider, you
can select between a snapshot (depending on the provider
configuration, this is a standard snapshot or vsnap) or a MirrorClone
snapshot. See "P6000 EVA specific options " on page 38.
o
With the HPE P9000 XP Disk Array Family hardware provider, a
snapshot is created.
The replica type is determined by the VSS hardware provider configuration.
Not available for instant recovery.
A provider may support one or both types. If you select an unsupported replica type, the backup
will fail.
For more information on supported replica types, see "Zero downtime backup with the Data
Protector VSS integration" on page 28.
l
To mount the replicas on the backup system, select Mount the replica on the backup
system.
VSS transportable backup options
HPE Data Protector (9.08)
Page 35 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
l
To change the replica management and mount options, click on Settings to open the Replica
management or the Mount dialog box.
Note: Whenever you change an option on the page that affects these additional settings,
the buttons are marked with an asterisk (*). After you open the settings window and review
or modify the options, the asterisk is removed.
For descriptions of replica options, see "Replica management options " on page 38 and "P6000
EVA specific options " on page 38 or press F1.
For descriptions of mount options, see "Mount options" on page 39 or press F1.
Click Next.
6. In this page, the selection of your VSS client is displayed.
On Windows Server 2008, you can specify the User and group/domain options. For information
on these options, press F1.
Click Next.
7. Select the backup objects you want to back up. Make sure that in case of instant-recovery
enabled sessions, you select all objects (for example all storage groups, all virtual machines ...)
residing on a specific source volume that will be backed up.
You can specify a full client backup by selecting the top-level item (the name of the client), a single
writer backup, or a writer’s component backup by selecting a lower-level item.
HPE Data Protector (9.08)
Page 36 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
If a full client backup is selected, Data Protector checks which writers exist on the client and
backs up all of them at backup time.
If a writer requires all of its components to be backed up, lower-level items are automatically
selected. If you select such a writer for backup, all its components will be backed up. If a writer
has no components to be backed up, it is not displayed in the list of writers, and is not backed up
when the full client is selected.
The Filesystem item displays all mounted disks. If another disk is mounted to a directory on a
disk, the parent disk name is displayed twice. The first name represents the parent disk name (for
example c:), while the second name represents the container for the mountpoint (for example
c:\mnt\1). To select the mounted disk, select the container for the mountpoint.
8. For backup to tape, select the devices you want to use for the backup to tape. Click Properties to
set the device concurrency, media pool, and preallocation policy. For more information on these
options, click Help. If you do not select a device, only backup to disk will be available.
If you do not configure devices and do not select the option Track the replica for instant
recovery, you will not be able to restore the data with Data Protector.
In case of backup to tape, you can also specify whether you want to create additional copies
(mirrors) of the backup during the backup session. Specify the desired number of mirrors by
clicking the Add mirror and Remove mirror buttons. Select separate devices for the backup and
for each mirror.
For detailed information on the object mirror functionality, see the HPE Data Protector Help.
9. Following the wizard, select the backup options and schedule your backup.
Tip: If you are not sure about selecting the backup options, keep the default values.
For details about the options common to all Data Protector backup specifications, see the HPE
Data Protector Help.
10. Once you have defined all backup options and the schedule, you need to name and save the
newly-created backup specification. You have now completed the creation of a Microsoft Volume
Shadow Copy Writers backup specification.
11. You can review the newly-created and saved backup specification in the Backup context, under
the specified group of backup specifications.
12. You can run backup using one of the following methods:
l
l
Schedule the backup of an existing Microsoft Volume Shadow Copy Writers backup
specification using the Data Protector Scheduler.
Start an interactive backup of an existing Microsoft Volume Shadow Copy Writers backup
specification.
Backup options
VSS-specific backup options
Option
Description
Pre-exec
Specify a command that will be started by vssbar.exe on the application
HPE Data Protector (9.08)
Page 37 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
Option
Description
system directly before replica creation. Do not use double quotes. Type only
the name of the command, not the pathname. The command must reside in the
default Data Protector commands directory.
Post-exec
Specify a command that will be started by vssbar.exe on the application
system directly after the replica creation. Do not use double quotes. Type only
the name of the command, not the pathname. The command must reside in the
default Data Protector commands directory.
ZDB options
Replica management options
Option
Description
Configuration
check mode
If you have selected Track the replica for instant recovery, specify the
Configuration check mode. Configuration check applies to the disk backups
that are to be used for instant recovery and does not apply to the tape backup.
For information on the options, see "Configuration check" on page 32.
Number of
replicas rotated
During ZDB sessions, Data Protector creates a new replica and leaves it on
the array until the specified Number of replicas rotated is reached (specify if
you selected Keep the replica after the backup). After that, the oldest replica
is deleted and a new one created.
Default: 3
The maximum number for vsnaps and standard snapshots is limited by the
P6000 EVA storage system. Data Protector does not limit the number of
replicas rotated, but the session fails if the limit is exceeded.
P6000 EVA specific options
Option
Description
Replica sub
type
Available if Keep the replica after the backup is selected.
Sets the replica subtype. Depending on the Replica type you selected, the
following subtypes can be selected:
l
Plex (Clone/Mirror):
l
l
HPE Data Protector (9.08)
SnapClone
Differential (Snapshot):
l
Snapshot
l
MirrorClone Snapshot
Page 38 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
Option
Description
Disabled if Replica type is set to Default.
Wait for data
copy to
complete
Available if Replica subtype is set to SnapClone.
Attempt to
minimize wait
This option resumes the backup process after the copy process has finished if
it is finished before the specified time.
During VSS backup session, a target storage allocation for the snapclone
replica is created first. At that point, the replica is only virtual but immediately
available for use. However, in the background, copy process runs for some
time to copy all data from the source to the target storage. To avoid the
slowdown of the copy process caused by further actions, for example
Exchange integrity check process, use this option. The backup resumes after
the time specified in Wait up to n minutes (default: 60) is reached even if the
copy process is not finished.
Selected by default if Wait for data copy to complete is selected.
Note: During the creation of the backup specification, the disk array type is not determined yet by
Data Protector and the P6000 EVA options are available regardless of the disk array used. If a
different disk array is used, the options are ignored when the backup is performed.
Mount options
Option
Description
Mount the
replica
Select this option to mount the replica on backup system. The mount path
depends on the Root of the mount point on the backup system and Add
directories to the mount path options.
Enable backup
system in
read/write
mode
Select this option to enable read/write access to the mounted replica's disks on
the backup system. Note that instant recovery of disks that have been
activated in read/write mode is not recommended, since setting the write
permission allows the disks' data to be changed after the backup time. Default:
not selected if Track the replica for instant recovery is on, or selected if only
Keep the replica after the backup is on.
Root of the
mount point on
the backup
system
Specifies the root directory under which the filesystems from the replica will be
mounted. Where exactly the filesystems are mounted depends on how you
specify Add directories to the mount path. The default mount point is
c:\mnt.
Add directories
to the mount
path
This option allows discrete control over the created mount points. When
Session ID is used in path creation, this guarantees unique mount points. The
options define which subdirectories will be created in the directory specified
with the Root of the mount path on the backup system option.
HPE Data Protector (9.08)
Page 39 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
Option
Description
Example:
Root directory: c:\mnt
Application system: app.comp.com
Backup session ID: 2008–02–22–4
Mount path on the application system: E:\disk1
If Hostname is selected: c:\mnt\app.comp.com\E\disk1
If Hostname + session ID is selected: c:\mnt\app.comp.com\2008–02–22–
4\E\disk1
If Session ID is selected: c:\mnt\2008–02–22–4\E\disk1
If Session ID + hostname is selected: c:\mnt\2008–02–22–
4\app.comp.com\E\disk1
Default: Hostname + session ID.
Automatically
dismount the
filesystems at
destination
mount points
If the mount points are in use (for example, a previous session may be
mounted) and this option is selected, Data Protector will try to dismount the
mounted filesystems. If the option is not selected and the mount points are in
use, or if the option is selected and the dismounting fails, the session will fail.
Default: not selected.
Scheduling backup sessions
For more detailed information on scheduling, see the HPE Data Protector Help index: “scheduled
backups”.
To schedule a Microsoft Volume Shadow Copy Writers backup specification, perform the following
steps in the Data Protector GUI:
1. In the HPE Data Protector Manager, switch to the Backup context.
2. In the Scoping Pane, expand Backup, then Backup Specifications. Click MS Volume Shadow
Copy Writers.
A list of available backup specifications is displayed in the Results Area.
3. Double-click the backup specification you want to schedule and click the Schedule tab to open
the Schedule property page.
4. In the Schedule property page, select a date in the calendar and click Add to open the Schedule
Backup dialog box.
5. Specify Recurring, Time options, Recurring options, and Session options.
Note that the backup type for ZDB sessions that are to be used for instant recovery is set to Full.
In the case of a ZDB-to-disk or a ZDB-to-disk+tape session, specify the Split mirror/snapshot
backup option.
HPE Data Protector (9.08)
Page 40 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
Scheduling a backup
Scheduling a ZDB backup
HPE Data Protector (9.08)
Page 41 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
6. Click OK to return to the Schedule property page.
7. Click Apply to save the changes.
Starting backup sessions
An interactive backup can be started using the Data Protector GUI by following these steps:
HPE Data Protector (9.08)
Page 42 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 4: Backup
1. In the HPE Data Protector Manager, switch to the Backup context.
2. In the Scoping Pane, expand Backup; then expand the Backup Specifications and the MS
Volume Shadow Copy Writers items.
3. Right-click the backup specification you want to use, and then select Start Backup from the popup menu.
The Start Backup dialog box appears.
Select the backup type and the network load (High, Medium, or Low).
Note that the backup type for ZDB sessions that are to be used for instant recovery is set to Full.
In the case of a ZDB-to-disk or a ZDB-to-disk+tape session, specify the Split mirror/snapshot
backup option.
For a description of network load, see the HPE Data Protector Help.
4. Click OK. Upon successful completion of the backup session, a Session Completed
Successfully message appears.
HPE Data Protector (9.08)
Page 43 of 119
Chapter 5: Restore
You can restore the Data Protector Microsoft Volume Shadow Copy Service integration backup objects using
the Data Protector GUI.
Data Protector offers two methods for restoring the writers:
l
From backup media to the application system on LAN (standard restore). See "Standard restore" below.
l
Using the instant recovery functionality. See "Instant recovery" on page 48.
Standard restore
Limitation for custom restore methods
l
Data Protector Microsoft VSS integration does not automatically provide any restore method for writers
requesting custom restore. If a writer specifies custom restore method, it is only possible to restore the
writer's data as plain files using the Data Protector file restore functionality. You can use the Restore
Into option to specify an alternate restore path for these plain files. You can then perform the custom
restore from these plain files manually. For information on writer’s custom restore, see the writer's
documentation.
Restore modes
Data Protector offers two restore modes:
l
Component restore using the VSS service
l
File restore using the Data Mover Agent (DMA) instead of VSS
By default, Data Protector restores writer components using the VSS service.
Component restore
When restoring writer components, Data Protector works with the VSS service to ensure data consistency:
1. Data Protector first restores the metadata, which was collected during the backup. Then it uses the
metadata to identify the backup components and determine the restore method.
2. Data Protector restores the data from the backup media to the locations specified in the backup
metadata, following the writers' instructions regarding any additional checking or processing.
3. After a successful restore from the backup media, Data Protector notifies the coordinator and the writers
can now access the newly-restored data and start the internal processing, for example database
recovery.
File restore
For a successful restore of a writer component, all files comprising this component must be restored. If a
restore of a single file fails, the restore of the a whole component fails. Data Protector offers an additional
restore mode for restoring single files that does not use the Microsoft Volume Shadow Copy Service, thus
HPE Data Protector (9.08)
Page 44 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
solving this problem. This mode can also be used for restoring to systems that do not support VSS or
do not have a VSS writer installed.
When restoring files or a group of files, DMA is started and the files are restored using the standard
Data Protector filesystem restore procedure.
As the file restore mode does not utilize VSS services, additional tasks that are performed after a
component restore – such as database recovery – are not performed and your application data may be
left in an inconsistent state, requiring additional manual procedures before the application is recovered.
Restoring using the GUI
The procedure below shows how to restore Microsoft VSS components using the Data Protector GUI.
Some writers require custom restore procedures and/or have specific limitations. See also the
appropriate sections in "Writer specifics" on page 62.
To restore Microsoft VSS objects using the Data Protector GUI, proceed as follows:
1. In the Data Protector GUI, switch to the Restore context.
2. Expand Restore Objects, expand MS Volume Shadow Copy Writers, expand the client from
which you want to restore the data, and then click MS Volume Shadow Copy Writers. In the
Results Area, a list of writers, which were backed up on this client, is displayed.
3. Select the Restore mode:
l
Restore components
With this option selected, whole components are restored using the Volume Shadow Copy
Service. Individual files cannot be selected for restore.
l
Restore files to temporary location
With this option, you can select individual files or a group of files that were backed up using the
selected writer. The files are restored using the Data Mover Agent and not the Volume Shadow
Copy Service.
4. In the Results Area, select the writers or writers’ components (for component restore) or files or a
group of files (for file restore mode).
Restore objects
HPE Data Protector (9.08)
Page 45 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
You can select the top-level item (full writer restore) or only specific components. If you select a
full writer restore, but some components of this writer were not backed up in the same session, the
unavailable components are shaded and cannot be selected. To select the version (the date of a
backup), right-click the object name and click Properties. The last available backup version is
selected by default, however, you can select a different version from the drop-down list.
For application-specific options, see "Writer specifics" on page 62.
5. In the Options property page, select the MS Volume Shadow Copy specific restore options. See
"Restore options" below.
6. In the Devices and Media property pages, the devices and media for restore are automatically
selected.
Note that you can change the device used for the restore. Therefore, you have the possibility of
using a different device for a restore than the one that was used for the backup. See the HPE Data
Protector Help index: “selecting devices for restore”.
7. Click Restore. Review your selection, and then click Finish to start a restore session.
The restore session messages are displayed in the Results Area.
8. If you are restoring a VSS writer that requires a custom restore, continue manually, using the
writer-specific methods, if it is provided by a writer. See the writer's documentation.
Restore options
The following restore options are specific to the Data Protector Microsoft Volume Shadow Copy
Service integration.
HPE Data Protector (9.08)
Page 46 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Note: Do not use these options for Microsoft Exchange Server 2007 writer since there are other
Exchange Server 2007 specific options provided for restore to another location. See the Microsoft
Exchange Server 2007 writer specifics.
Restore Options
Restore Options
Description
Restore to another
client
By default, the components or files are restored to the client from which the
application data was backed up. However, you may restore the data to
another VSS client if you specify the Restore to another client option. The
new target
Microsoft VSS client must be a part of the Data Protector cell. For
component restore, it must also run on the same platform and have the MS
Volume Shadow Copy Integration software component installed. For file
restore, the MS Volume Shadow Copy Integration software component is
not required.
Restore into the
following directory
By default, you restore the data to the same directory from which it was
backed up (it can be on the original client or on some other client which you
selected).
However, if you specify the Restore into the following directory option,
your data will be restored to another directory. When defining the restore
location, you can specify the path to the directory where you want to restore
your data.
Restoring using the CLI
Use the omnidbvss -get session SessionKey command to get details about the session that you
want to use for restore.
For example:
omnidbvss -get session_persistent 2011/01/26-2:computer1
==========================================================
Session ID:
2011/01/26-2
Barlist Name:
SQL_NEW_DB_SIMPLE
Bar Hostname:
computer1.company.com
Backup Type:
FULL
Instant Restore:
FALSE
Disk-Only:
FALSE
Component Name
==========================================================
[0]
/SqlServerWriter(SQL Server 2008:SQLWriter)/COMPUTER1/New_DB
The objects that were backed up in this session are listed under Component Name. See the omnidbvss
reference page in the HPE Data Protector Command Line Interface Reference for details.
Execute the omnir command:
HPE Data Protector (9.08)
Page 47 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
omnir -vss –barhost ClientName -session BackupID -tree TreeName1 [-tree
TreeName2...] -conf_check {strict|non-strict|disabled}
See the omnir reference page in the HPE Data Protector Command Line Interface Reference for
details.
Note: BackupID is a point in time. All objects (backup data) created in a backup session have the
same backup ID, which is the same as the session ID of the backup session.
Mirrored objects and objects created in an object copy session have the same backup ID as
objects created in the original backup session. Suppose the media set created in the original
backup session no longer exists, but the media set created in an object copy session still exists.
To restore the objects, you must specify the session ID of the original backup session (that is, the
backup ID) and not the session ID of the object copy session.
The omnir syntax does not let you specify from which object copy to restore if several copies of
the same object exist. This is only possible using the Data Protector GUI by setting the media
allocation priority list.
To recover the components from the above example, execute:
omnir -vss –barhost server1.company.com -session 2010/1/12-09 -tree /Microsoft
Exchange Writer(Exchange Information Store)/Microsoft Information Store/Store1/Logs
–tree /Microsoft Exchange Writer(Exchange Information Store)/Microsoft Information
Store/Store1/File -conf_check disabled
Instant recovery
After ZDB to disk+tape or ZDB to disk, restore can be performed using instant recovery. After ZDB to
disk, instant recovery is the only possible restore method. Backup session information is saved in the
IDB, and the array-specific information required for instant recovery is saved in the VSS database
(VSSDB).
Instant recovery consists of two phases:
l
l
An instant recovery session is started and the relevant data files are restored back to the system.
A writer performs a recovery of the application data, which is based on the restored data. For some
writers, database recovery is performed automatically by the database engine. See the writer
specifics for details.
During instant recovery, the whole replica is restored. This means not only are the originally selected
backup objects restored, but the complete content of all the volume groups that contained them. After
the instant recovery, the content is returned to the state when the replica was created.
For detailed information on instant recovery concepts, see the HPE Data Protector Concepts Guide
and HPE Data Protector Zero Downtime Backup Administrator's Guide.
You can selectively restore separate writer’s components only if they reside on separate target
volumes. Instant recovery of separate components also requires that the target volumes with
components data should not contain any other data. If you do not select all of the writer’s components
residing on the target volumes to be restored, the restore fails.
Before starting an instant recovery, Data Protector checks:
HPE Data Protector (9.08)
Page 48 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
l
l
Whether the data that is required for the components restore is available. If the check fails, the
instant recovery session also fails, ensuring that no data that does not belong to the components is
lost.
If any changes that are not controlled by Data Protector happened in a disk array environment (for
example, the target volumes to be restored are presented to some other system as it is recorded in
the VSSDB, or the source volumes to be replaced are not presented on the application system). If
Data Protector detects such changes (by comparing the state of the environment with data in the
VSSDB), the instant recovery session aborts. However, you can force the session to continue by
setting the OB2VSS_IGNORE_BACKUP_DISK_CHANGES and OB2VSS_IGNORE_SOURCE_DISK_CHANGES
omnirc options to 1.
You can perform an instant recovery using either the Volume Shadow Copy or Virtual Disk services or
the Data Protector disk array integration agents:
l
Using Microsoft Virtual Disk Service (VDS)
The instant recovery is initiated by Data Protector and performed by the VDS provider. Information
about the replica is deleted from the VSSDB, therefore it is not possible to perform another instant
recovery using that replica.
l
Using Microsoft Volume Shadow Copy Service LUN resync
The instant recovery is performed by the VSS hardware provider. The actual instant recovery
method depends on the disk array and VSS hardware provider settings. The VSS LUN resync
functionality must be supported by the operating system (Microsoft Windows Server 2008 R2).
For a list of hardware providers that support this type of recovery, see the latest support matrices on
the web https://softwaresupport.hpe.com/.
l
Using a Data Protector disk array integration agent (HPE P4000 VSS Agent, HPE P6000 / HPE
3PAR SMI-S Agent, HPE P9000 XP Agent, or HPE 3PAR VSS Agent)
The instant recovery is performed by the appropriate Data Protector disk array agent.
The available instant recovery methods depend on the disk array, the way the instant recovery is
performed (either by using one of the hardware providers or Data Protector disk array agents), and on
the type of the backup.
Instant recovery methods
Disk array technologies support several methods of instant recovery:
l
Switch of disks
l
Copy of replica data
l
Restore of snapshot data
l
Resync of volumes
Some of the methods also enable you to keep the source volumes while with others, the source
volumes are not retained. For details, see the description of each method. For a complete overview of
all available instant recovery methods for a specific disk array, see "Supported instant recovery
methods for HPE P4000 SAN Solutions" on page 53, "Supported instant recovery methods with HPE
P6000 EVA Disk Array Family" on page 53, " Supported instant recovery methods for HPE P9000 XP
Disk Array Family" on page 54, or "Supported instant recovery methods for HPE 3PAR StoreServ
Storage " on page 56.
HPE Data Protector (9.08)
Page 49 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Switch of disks
With this method, the source volumes are unpresented and the target volumes (replica) are presented
in the place of the source volume. You can select to retain the old source volumes. However, you
cannot retain the replica and perform another instant recovery from this replica.
If the restore session fails or is aborted, the VSS integration reverts the application and backup
environment back to the state in which they were before the instant recovery.
Advantages:
l
Instant recovery is very fast.
l
The original source volume can be retained after instant recovery.
l
By selecting the Restore using Microsoft Virtual Disk Service option, you can restore your data
without using HPE P6000 / HPE 3PAR SMI-S Agent or HPE P9000 XP Agent.
Disadvantages:
l
l
It is not possible to perform another instant recovery using the same replica.
This method of instant recovery changes the physical location of the application data, since, after
instant recovery, a replica becomes the source volume. The application starts running on the
physical disks that were used for backup. If the selected replica and its source volume belong to
separate disk groups, the two disk groups are also switched during instant recovery.
To use this method of instant recovery, in the Data Protector GUI, select the Restore using Microsoft
Virtual Disk Service option.
Copy of replica data
Copy of replica data with the source volume retained
With this method, a new copy from a replica is created in the same disk group in which the source
volumes reside. When the copy is finished, a switch of disks between the source and the new copy is
performed. The old source volumes are retained.
If the restore session fails or is aborted, the VSS integration reverts the application and backup
environment back to the state in which they were before the instant recovery.
The replica is automatically retained thus you can perform another restore from this backup.
To delete the replica, you can choose to manually remove it using the omnidbvss command or wait until
it is deleted automatically due to replica rotation.
Advantages:
l
The original source volumes are retained after restore.
l
Disk group of the source volume is not changed after instant recovery.
l
Another restore from the same backup is possible.
Disadvantages:
l
Restore is not as fast as with the “switch of disks” method.
l
Physical location of the application production data changes.
HPE Data Protector (9.08)
Page 50 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
To perform this type of restore, select the Copy replica data to the source location option and leave
selected the Retain source for forensics option in the Data Protector GUI.
Copy of replica data with the source volume not retained
With this method, the source volumes are directly overwritten by the replica by copying data from the
target volumes back to the source volumes. The old source volumes are not retained and if the restore
session fails after the copy process has already started, the original application data residing on the
source volumes is lost. If you try to abort the session at this point, the abort operation is rejected and
the session continues.
The replica is automatically retained thus you can perform another restore from this backup.
To delete the replica, you can choose to manually remove it using the omnidbvss command or wait until
it is deleted automatically due replica rotation.
Advantages:
l
Disk group of the source volume is not changed after instant recovery.
l
Physical location of the application production data remains the same.
l
Another restore from the same backup is possible.
Disadvantages:
l
Restore is not as fast as with the “switch of disks” method.
l
The original source volumes are lost during restore.
To perform this type of restore, select the Copy replica data to the source location option and clear
the Retain source for forensics option in the Data Protector GUI.
Restore snapshot data to the source volume
Restore snapshot data to the source volume with the source volume
retained
With this method, the data from the snapshots is copied back to the original volumes.
If the restore session fails or is aborted, the VSS integration reverts the application and backup
environment back to the state in which they were before the instant recovery.
The replica is automatically retained thus you can perform another restore from this backup.
To delete the replica, you can choose to manually remove it using the omnidbvss command or wait until
it is deleted automatically due replica rotation.
Advantages:
l
The original source volumes are retained after restore.
l
Disk group of the source volume is not changed after instant recovery.
l
Another restore from the same backup is possible.
l
Because a snapshot contains only differential data, the amount of copied data is usually lower
compared to copying a replica that is a full copy of its source (for example a snapclone).
Disadvantages:
HPE Data Protector (9.08)
Page 51 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
l
Restore is not as fast as with the "switch of disks" method.
To perform this type of restore, select the Copy replica data to the source location option and leave
selected the Retain source for forensics option in the Data Protector GUI.
Restore snapshot data to the source location with the source volume not
retained
With this method, a snapshot of the source volume is made first and then the data from the snapshots
is copied back to the original volumes.
Advantages:
l
Disk group of the source volume is not changed after instant recovery.
l
Physical location of the application production data remains the same.
l
Another restore from the same backup is possible.
l
Because a snapshot contains only differential data, the amount of copied data is usually lower
compared to copying a replica that is a full copy of its source (for example a snapclone).
Disadvantages:
l
Restore is not as fast as with the "switch of disks" method.
l
The original source volumes are lost during restore.
To perform this type of restore, select the Restore snapshot data to the source location option and
clear the Retain source for forensics option in the Data Protector GUI.
Resync of volumes
With this method, the target volumes are resynchronized with the original source volumes.
With this type of instant recovery, HPE P9000 XP Agent (SSEA) synchronizes a source volume (PVOL) with its target volume (S-VOL) and then splits the pair during the instant recovery session.
Depending on the selected resync type, SSEA waits for re-synchronization or copy process to
complete (normal restore, selected with Resync replica data to the source volume), before making
the source volume available. Alternatively, the source volume can be made immediately available while
the re-synchronization or copy process is running in the background (quick restore, selected with
Quick-resync replica data to the source volume).
If you try to abort the session after the copy process has started, the abort operation is rejected and the
session continues.
The original source volumes are not retained after instant recovery. If the session fails after the copy
process has already started, the application data residing on the original source volumes is lost.
Limitations
l
l
Instant recovery cannot be performed after a point-in-time recovery. To be able to perform instant
recovery, you first need to run backup with the instant recovery options set.
The number of replicas available for instant recovery is limited by Number of replicas rotated,
which sets the size of the replica set. You can view these replicas in the GUI, in the Instant
HPE Data Protector (9.08)
Page 52 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Recovery context by expanding Restore Sessions. Replicas are identified by the backup
specification name and the session ID. Other information, such as time when a particular replica
was created, is also provided. Alternately, you can use the omnidbvss command to list sessions.
For information about omnidbvss, see the HPE Data Protector Command Line Interface Reference.
l
l
Concurrent instant recovery sessions utilizing the same backup system or the same application
system are not supported. Restores utilizing the same backup system or the same application
system must be performed serially.
For instant recovery of Exchange Server 2007 data to a different location, the Microsoft Virtual Disk
Service, HPE P4000 SAN Solutions, HPE P6000 EVA Disk Array Family, HPE P9000 XP Disk
Array Family, and HPE 3PAR StoreServ Storage restore options are not available.
For additional writer specific limitations, see "Writer specifics" on page 62.
HPE P4000 SAN Solutions considerations
l
The following table lists instant recovery methods available with the HPE P4000 SAN Solutions
Supported instant recovery methods for HPE P4000 SAN Solutions
Instant recovery method
Restore snapshot data to the source
volume
l
l
HPE P4000
VSS Agent
with the source volume retained
No
with source volume not retained
Yes
HPE P4000 SAN Solutions replicas depend on previous replicas in a chain. This means that in order
to restore a particular replica, newer replicas are deleted. When performing instant recovery, try to
use the newest replica possible and then gradually restore other replicas to the desired point in time,
to avoid accidental replica loss.
Any newer replicas created outside of Data Protector will prevent instant recovery from a session
that contains the volume which was replicated outside of Data Protector.
HPE P6000 EVA Array considerations
l
The following table lists instant recovery methods available with the HPE P6000 EVA Disk Array
Family:
Supported instant recovery methods with HPE P6000 EVA Disk Array Family
HPE Data Protector (9.08)
Page 53 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Instant recovery method
VDS
hardware
provider
HPE
P6000 /
HPE 3PAR
SMI-S
Agent
Switch of disks
Yes1
Yes2
with the source retained
No
Yes
with the source not retained
No
Yes
with the source volume retained
No
Yes
with source volume not retained
No
Yes
Copy of replica data
Restore snapshot data to the
source volume
l
If you restore using P6000 EVA Array, before instant recovery, it is recommended to check if the
SMI-S CIMOMs were configured properly in the Data Protector cell by executing the omnidbsmis ompasswd -check [-host ClientName] command.
This command performs a health check of you environment, which may help identify such potential
problems as wrong user name or password provided, a broken network connection, a DNS
resolution problem, and so on.
l
l
l
l
l
After instant recovery, restored filesystems are mounted to the same mount points or drive letters as
they were at the backup time. If these mount points or drive letters have other filesystems mounted,
these filesystems are automatically dismounted before instant recovery, and the restored
filesystems are mounted afterwards.
You can restore snapshots or MirrorClone snapshots only to the source volume.
On Windows Server 2008 systems, if the VDS P6000 EVA Array hardware provider was installed at
the time of the backup, the VDS P6000 EVA Array hardware provider is required for instant recovery.
On Windows Server 2003 systems, the VDS P6000 EVA Array hardware provider is always required
for instant recovery.
If the VDS P6000 EVA Array hardware provider is not used, Data Protector can perform fewer
checks during instant recovery which may decrease the ability to revert changes in case of failure.
If using Command View 9.2.1, instant recovery of a snapshot or MirrorClone snapshot can only be
performed after the cache on the Command View server has been refreshed (triggered by default
every 30 minutes).
HPE P9000 XP Array considerations
l
The following table lists instant recovery methods available with the HPE P9000 XP Array:
Supported instant recovery methods for HPE P9000 XP Disk Array Family
1Available if the backup was performed using snapclone as subtype.
2Available if the backup was performed using snapclone as subtype.
HPE Data Protector (9.08)
Page 54 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Instant recovery method
VDS
hardware
provider
HPE
P9000 XP
Agent
Switch of disks
Yes1
No
Resync of volumes
No
Yes2
with the source volume retained
No
No
with source volume not retained
No
Yes3
Restore snapshot data to the
source volume
l
l
Instant recovery depends on the P9000 XP Array VSS hardware provider mode which was used
during backup. If the VSS compliant mode was used, you can restore your data only using VDS. If
the resync mode was used, you can restore your data only using SSEA.
Changing the selected P9000 XP Array VSS hardware provider mode between different backup
sessions using the same backup specification is not recommended. If the mode is changed while
the replica set rotation count is set to more than 1, instant recovery will fail in the following
situations:
l
If you perform a backup in one mode and then the same backup in the other mode, restore of the
backup performed in the VSS compliant mode (the “switch of disks” restore) will fail because the
relationship between the S-VOL and its P-VOL will be detected from the other backup performed
in the resync mode. This pair relationship should not be removed during restore and thus a switch
of the source volume with the replica cannot be performed.
To restore such a backup, perform one of the following prior to restore:
l
o
Using the omnidbvss command, manually remove S-VOLs that were created during backups
in resync mode.
o
Set the option OB2VSS_FORCE_INSTANT_RECOVERY in the omnirc file and enable the restore
option Retain source for forensics in the GUI. When this option is selected, Data Protector
preserves the pair relationship between the S-VOL and its P-VOL created in the backup with
resync mode.
If you perform restore using re-synchronization (the "copy of replica data" restore) and the
production source volume (P-VOL) is not presented on the system, the restore session is
aborted. This may happen after at least two backup sessions are run using the same backup
specification in different modes and you run a “switch of disks” restore before a "copy of replica
data" restore.
Under such circumstances, perform one of the following prior to restore:
l
o
On the application system, manually present the P-VOL.
o
Set the option OB2VSS_FORCE_INSTANT_RECOVERY in the omnirc file.
In VSS compliant mode, the VDS P9000 XP Array hardware provider is required for instant recovery,
1hardware provider in the VSS compliant mode
2hardware provider in the resync mode
3hardware provider in the resync mode
HPE Data Protector (9.08)
Page 55 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
even if the backup was performed without it. In such a case, install the VDS P9000 XP Array
hardware provider and execute the omnidbvss -resolve command before you attempt an instant
recovery:
l
l
To resolve the source volumes on the application system: omnidbvss -resolve -apphost
ApplicationClient
To resolve the target volumes created in the backup session: omnidbvss -resolve -session
SessionID
The following restore considerations apply to restore using the HPE P9000 XP Agent only:
l
l
You cannot start another instant recovery session using the same disk on the application system at
the same time. A session can be started only after the preceding session completes
synchronization.
After instant recovery, the restored filesystems are mounted to the drive letters as were used at the
time when backup was run. If these drive letters have other filesystems mounted, these filesystems
are automatically dismounted before instant recovery, and the restored filesystems are mounted
afterwards.
HPE 3PAR StoreServ Storage considerations
l
The following table lists instant recovery methods available with the HPE 3PAR StoreServ Storage:
Supported instant recovery methods for HPE 3PAR StoreServ Storage
Instant recovery method
Restore snapshot data to the source
volume
l
HPE 3PAR
VSS Agent
with the source volume retained
No
with source volume not retained
Yes
HPE 3PAR StoreServ Storage replicas do not depend on previous replicas in a chain.
Instant recovery procedure
Limitations
l
l
Instant recovery of Cluster Shared Volumes is not supported.
Snapshots can be restored only to the original LUN (the LUN from where the replica was created)
and to the same client.
Procedure
1. In the HPE Data Protector Manager, switch to the Instant Recovery context.
2. In the Scoping Pane, expand MS Volume Shadow Copy Writers under Restore Objects,
expand the backup specification from which you want to restore, and click the backup session
HPE Data Protector (9.08)
Page 56 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
from which you want to restore.
Selecting writers components for instant recovery
3. In the Source property page, specify writers and/or components for recovery.
Select the configuration check mode. For more information, see "Configuration check" on page 32.
To perform additional steps after the instant recovery before recovering the database, select the
No recovery option. Data Protector will not perform a database recovery and you can finish the
recovery steps (such as applying transaction logs) later on manually.
If the option is not selected, all recovery steps are finished. In such a case, additional tasks
cannot be performed. This option depends on the application writer that is being recovered._
For any additional application-specific options, see "Writer specifics" on page 62.
4. Under the Options tab, select the instant recovery type:
l
Restore using Microsoft Virtual Disk Service
l
Restore using Microsoft Volume Shadow Copy Service LUN resync
l
Restore using the disk array agent
See "Instant recovery methods" on page 49 for details on which type is supported with which disk
array.
HPE Data Protector (9.08)
Page 57 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
For instant recovery using the VDS hardware provider or Disk Agent, to retain the source volume,
select Retain source for forensics.
When a recovery using a disk array agent is selected, additional options are available.
l
Restore using HPE P4000
The restore method: Restore snapshot data to the source volume.
For details on these methods see "Instant recovery methods" on page 49.
l
Restore using HPE P6000 EVA SMI-S
The restore method: Copy replica data to the source volume, Restore snapshot data to
the source volume, or Switch the replica. For details on these methods see "Instant
recovery methods" on page 49.
If you select Copy replica data to the source volume, a full copy of the replica is created in
the source storage (the copy is then switched with the source volume). A virtual copy is
created and immediately available for use. However, in the background, the copy process runs
for some time to copy all data from the replica to the source storage. To avoid the slowdown of
the copy process caused by application use or the integrity check process (Exchange Server),
use this option. The restore resumes after the time specified in Wait up to n minutes (default:
60) is reached (even if the copy process is not finished) or after the copy process has finished if
it is finished before the specified time.
l
Restore using HPE P9000 XP
The restore method: Resync replica data to the source volume, Quick-Resync replica
data to the source volume (internal swap), or Restore snapshot data to the source
volume.
For details on these methods see "Instant recovery methods" on page 49.
l
Restore using HPE P10000 3PAR
The restore method: Restore snapshot data to the source volume.
For details on these methods see "Instant recovery methods" on page 49.
For details, press F1.
Selecting instant recovery options (P4000 SAN Solutions integration)
HPE Data Protector (9.08)
Page 58 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Selecting instant recovery options (P6000 EVA Array integration)
Selecting instant recovery options (P9000 XP Array integration, split mirror)
HPE Data Protector (9.08)
Page 59 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
Selecting instant recovery options (P9000 XP Array integration, snapshot)
Selecting instant recovery options (3PAR StoreServ integration)
HPE Data Protector (9.08)
Page 60 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 5: Restore
5. Click Restore.
6. Perform any additional writer specific steps. See "Writer specifics" on page 62.
Clusters
To perform an instant recovery in the cluster environment, use the above instant recovery procedure.
When selecting the client name, use virtual client names instead of physical systems. For details on
performing an instant recovery in cluster configurations, see the HPE Data Protector Zero Downtime
Backup Administrator's Guide.
HPE Data Protector (9.08)
Page 61 of 119
Chapter 6: Writer specifics
This chapter provides specific information about VSS writers, that you need to take into account before
backing up or restoring the writers.
VSS writers either come with the Windows operating system or with applications. For a complete list of
supported VSS writers and providers, see the latest support matrices at https://softwaresupport.hpe.com/.
The Data Protector Microsoft VSS integration does not provide any restore method for writers requesting a
custom restore. If a writer specifies a custom restore method, it is only possible to restore the writer's data as
plain files using the Data Protector functionality. You can perform the custom restore manually. For additional
information on the restore methods, see the writer's documentation.
Note: Writers requiring custom restore methods are by default not shown by Data Protector. The omnirc
option OB2VSS_SHOWALLWRITERS must be set to 1 for all writers to be displayed.
" Writer description " below provides a description of VSS writers.
Writer description
Writer name
Description
Restore method
Certificate
Authority Writer
This is a system writer, used to back up and restore
Certificate Authority (CA) Service database. This
service issues, revokes, and manages certificates
employed in public key-based cryptography
technologies.
Files are restored after a
system restart.
Cluster Service
Writer
This VSS writer uses a custom API, and is used to
back up and restore Cluster Service on Microsoft
Cluster Server (MSCS). The Cluster Service is a
component on Windows Server systems used to
control server cluster activities on cluster nodes. It is
fundamental to the operation of the cluster.
Custom restore method
COM+ REGDB
Writer
This VSS writer uses a custom API, and is used to
back up and restore COM+ Database Service. This
service provides automatic distribution of events to
subscribing COM+ components.
Custom restore method
DHCP Jet Writer
This is a system writer, used to back up and restore
DHCP Service database. DHCP Service provides
dynamic IP address assignment and network
configuration for Dynamic Host Configuration
Protocol (DHCP) clients.
Files are restored after a
system restart.
Event Log Writer
This is a system writer, used to back up and restore
Event Logs. Event Logs are files where the Windows
operating system saves information about events,
Files are restored after a
system restart.
HPE Data Protector (9.08)
Page 62 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Writer name
Description
Restore method
such as starting and stopping services or the logging
on and logging off of a user.
FRS Writer
This VSS writer uses a custom API, and is used to
back up and restore File Replication Service data.
File Replication Service is a multithreaded replication
engine that replicates system policies and logon
scripts stored in System Volume (SYSVOL). FRS can
also replicate data for Distributed File System (DFS),
copy and maintain shared files and folders on multiple
servers simultaneously.
Custom restore method
IIS Metabase
Writer
This is a system writer, used to back up and restore
Microsoft Internet Information Server (IIS). IIS is a
network file and application server that supports
multiple protocols. Primarily, IIS transmits
information in Hypertext Markup Language (HTML)
pages by using the Hypertext Transport Protocol
(HTTP).
Files are restored after a
system restart.
MSDE Writer
This is a writer used to back up and restore Microsoft
SQL Server 2000/2005. SQL Server is a database
management system that can respond to queries
from client machines formatted in the SQL language.
See "MSDE writer
specifics" on page 108.
Microsoft SQL
2005 Writer
This is a writer used to back up and restore Microsoft
SQL Server 2005. SQL Server is a database
management system that can respond to queries
from client machines formatted in the SQL language.
Standard VSS restore
and instant recovery.
Microsoft Data
Protection Manager
Writer
This is a writer used to back up and restore Microsoft
Data Protection Manager. Microsoft Data Protection
Manager is a server that creates and stores replicas
of clients and uses them for recovering the data on
clients
See "Microsoft Data
Protection Manager writer
specifics" on page 65.
Microsoft
Exchange Server
Writer
This is a writer used to back up and restore Microsoft
Exchange Server. Microsoft Exchange Server is a
mail and groupware server.
See "Microsoft Exchange
Server 2003 writer
specifics" on page 69,
"Microsoft Exchange
Server 2007 writer
specifics" on page 76, or
"Microsoft Exchange
Server 2010 writer
specifics" on page 92.
HPE Data Protector (9.08)
Page 63 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Writer name
Description
Restore method
Microsoft Virtual
Server 2005 Writer
This is a writer used to back up and restore Microsoft
Virtual Server 2005. Microsoft Virtual Server 2005 is a
virtualization platform for Microsoft Windows Server
systems. Data Protector supports live backup of
individual virtual machines and the Virtual Server
configuration, ensuring data consistency of the
backup and restore. Hardware providers are not
supported if a Virtual Server Machine is in online
mode; use a software provider or put the Virtual
Server Machine in offline mode. For details about
Virtual Server online and offline mode, see the
Microsoft Virtual Server documentation.
Standard VSS restore
and instant recovery.
Cluster configurations are not supported, only
individual nodes can be backed up.
Microsoft Hyper-V
Writer
This is a writer used to back up and restore Microsoft
Virtual Server 2008 Hyper-V configuration and
individual or all virtual machines running on the
server. Software and hardware providers are
supported during online and offline backup.
Standard VSS restore
and instant recovery. For
writer specifics, see
"Microsoft Hyper-V writer
specifics" on page 100.
Cluster-aware backups are not supported.
SharePoint
Services Writer
This is a reference writer used to back up and restore
Microsoft Office SharePoint Server 2007 (MOSS).
MOSS 2007 is an information portal used to connect
and exchange expertise between people and teams.
Only a single server configuration (farm) is supported.
Standard VSS restore.
For writer specifics, see
"Microsoft SharePoint
Services writer specifics"
on page 104.
The OSearch VSS writer and SPSearch VSS writer
are used by the reference writer to back up and
restore the index files of user and help content. They
must not be used for backup and restore.
NTDS Writer
This is a system writer used to back up and restore
Microsoft Active Directory on Windows Server
systems. Active Directory Service is a Windows
server directory service that enables you to manage
data structures distributed over a network. For
example, Active Directory Service stores information
about user accounts, passwords, phone numbers,
profiles, and installed services. It provides methods
for storing directory data and making this data
available to network users and administrators.
To restore Active
Directory, boot into
Directory restore mode.
Files will be restored if
they can be overwritten.
Registry Writer
This VSS writer uses a custom API, and is used to
Custom restore method
HPE Data Protector (9.08)
Page 64 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Writer name
Description
Restore method
back up and restore Windows Registry. Windows
Registry is a database repository of information
containing the Windows system configuration.
Remote Storage
Writer
This is a system writer used to back up and restore
Remote Storage Service (RSS). RSS is used to
automatically move infrequently accessed files from
local to remote storage. Remote files are recalled
automatically when the file is opened.
Files are restored after a
system restart.
Removable
Storage Manager
Writer
This is a system writer used to back up and restore
Removable Storage Manager Service. This service
manages removable media, drives, and libraries.
Files are restored after a
system restart.
System Writer
This is a system writer that backs up a specific set of
Windows dynamic link libraries (DLL).
Files are restored after a
system restart.
TermServLicencing
Writer
This is a system writer that backs up Windows
Terminal Services. These services provide a multisession environment that allows client systems to
access a virtual Windows desktop session and
Windows-based programs running on the server.
Files are restored after a
system restart.
WINS Jet Writer
This is a system writer, used to back up and restore
Windows Internet Name Service (WINS). WINS is a
dynamic replicated database service that can register
and resolve NetBIOS names to IP addresses used on
a TCP/IP network.
Files are restored after a
system restart.
WMI Writer
This is a system writer, used to back up and restore
Windows Management Instrumentation (WMI). WMI
is a unified management infrastructure in Windows
for monitoring system resources.
Files are restored after a
system restart.
Microsoft Data Protection Manager writer
specifics
Backup
Microsoft Data Protection Manager (DPM) is a server application that creates replicas of the clients,
synchronizes them through LAN, and stores these replicas as snapshots.
The Data Protection Manager writer is used to back up:
HPE Data Protector (9.08)
Page 65 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
l
The Data Protection Manager database and the Data Protection Manager Report database
l
The latest version of the DPM replicas
To ensure data consistency, schedule a DPM replica synchronization before starting a backup.
The DPM uses DPM snapshots for restore. These snapshots are not backed up. To be able to recreate
DPM snapshots you must manually schedule a backup of the replica each time after the DPM creates
a new replica.
Two backup types are supported:
l
Full (for the DPM databases and replicas)
l
Incremental (replicas only)
If you select unsupported backup types (Copy or Differential) when scheduling the backup, Data
Protector will abort the backup and display an error message.
Prerequisite
The MSDE writer (used for backing up the DPM databases) must be installed.
Limitations
l
l
Hardware providers are not supported with DPM.
If you start a backup while an incremental synchronization of the DPM replica is still in progress, the
backup gets corrupted, although Data Protector does not report any errors. In case of
synchronization with consistency check, Data Protector automatically aborts the backup session.
Selecting Microsoft Data Protection Manager database and replicas
HPE Data Protector (9.08)
Page 66 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Restore
When restoring the DPM writer, you can:
l
Restore the DPM server first and then use the DPM to restore clients.
In case of a disaster, when the entire DPM server is lost, perform a standard disaster recovery
procedure first and continue with restoring the DPM server. See "Restore the DPM server first"
below
l
Restore individual DPM clients directly, without using the DPM server (for example, if you cannot
restore the DPM server or if you want to avoid the additional step of recreating the DPM snapshot).
When restoring the DPM clients directly you can select between component restore and file restore
modes. See "Restore the DPM clients directly" on the next page.
Note: Although the Data Protection Manager databases could also be restored using the MSDE
writer, this method is not recommended, because DPM is not shut down automatically as with the
DPM writer. If you really need to use this writer, shut down the DPM server manually.
Limitations
l
Restore to another server is not supported by the Data Protection Manager writer.
l
Parallel restore to different clients is not supported.
Restore the DPM server first
1. Start the DPM administrator console and add disks to the storage pool so that you have enough
free space to restore the replicas.
Ensure that the DPM writer (service) is started.
2. Switch to the Data Protector Restore context. Expand Restore and Microsoft Volume Shadow
Copy Writers and select the client from which you want to restore the data.
3. In the Results Area, expand the DPM writer and select only the Data Protection Manager
databases.
Proceed as with general VSS writer restore. See "Standard restore" on page 44 for the general
VSS writer restore procedure.
4. Execute the DPM command DpmSync -Sync to reallocate replicas.
5. Switch back to the Data Protector Restore context, and select and restore the necessary
replicas.
Restoring the Microsoft Data Protection Manager client
HPE Data Protector (9.08)
Page 67 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
6. Use the DPM to restore individual clients.
The DPM console does not automatically check for new or restored snapshots. Before you can
start the restore of clients, you must use the Data Protection Manager to recreate a DPM
snapshot.
a. In the DPM console, open the Recovery context. Under the Browse tab, select the server,
right click on the restored replica, and select Create shadow copy now.
b. Select and restore the new snapshot to the client.
Restore the DPM clients directly
1. Switch to the Restore context. Expand Restore and Microsoft Volume Shadow Copy Writers
and select the client from which you want to restore the data.
2. Select the restore modes:
l
Restore Components
Use this mode only if the client to which you want to restore supports VSS, for example if you
restore to Windows Server 2003 clients.
You can restore only entire replicas.
l
Restore Files
The client does not need to support VSS and you can restore individual folders or files.
3. When selecting the DPM writer for restore, select only the Replica components. Do not select the
DPM database.
HPE Data Protector (9.08)
Page 68 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
4. Click the Options tab, and under Restore to another client enter the name of the target client.
Click Next.
5. Proceed as with general VSS writer restore. See "Standard restore" on page 44 for the general
VSS writer restore procedure.
Microsoft Exchange Server 2003 writer specifics
Backup
Backup types
The Microsoft Exchange Server Writer supports the following Microsoft Exchange backup types:
l
l
Full - backs up databases, transaction logs, and checkpoint files. The transaction logs are truncated.
Incremental - backs up the transaction logs to record changes since the last full or incremental
backup. The transaction logs are truncated.
This backup type is not available with the VSS hardware providers.
l
Differential - similar to incremental backup, but the transaction logs are not truncated.
This backup type is not available with the VSS hardware providers.
l
Copy - similar to full backup, with the difference that the transaction logs are not truncated. This
backup type is not intended for use in recovering failed systems.
Limitations
l
A combination of VSS snapshot backups and non-VSS backups (for example, stream incremental
backups) is not supported.
l
You can back up only the whole server or full storage groups. Single stores cannot be backed up.
l
Circular logging must be disabled.
l
l
With Exchange Server 2003, only one VSS backup session can be running on the same application
system at once. With Exchange Server 2007, only one VSS backup session backing up a specific
storage group can be running on the same application system at once. Consequently, if you start
such a backup session while another one is in progress, the latter waits until the first one finishes.
You must create a separate incremental backup specification for each backup specification that will
be used for instant recovery (has the Track the replica for instant recovery option enabled).
For example, you cannot create the backup specification BSpec1 for the storage group SGroup1
with snapshots as the backup type and the backup specification BSpec2 for the storage group
SGroup2 with snapclone as backup type and then use only one incremental backup specification
containing both storage groups.
Recommendations
l
With the HPE SMI-S P6000 EVA Array provider, an Exchange storage group should not reside on
more than four source volumes. However, the recommended configuration is that transaction logs
reside on one source volume and the stores on another source volume. This configuration enables
you to perform a rollforward recovery if only the stores are lost.
Create one backup specification for each storage group and schedule them in a row.
HPE Data Protector (9.08)
Page 69 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Rollforward recovery requirements
To ensure that a rollforward recovery from your backup is possible, consider the following:
Standard restore : Transaction logs must be backed up to enable the rollforward operation.
Instant recovery : Rollforward of a storage group can be performed with incremental or differential ZDB
to tape of the same objects as were backed up in instant recovery enabled sessions (ZDB to disk or
ZDB to disk + tape). If the logs are still available on disk, the restore of last full backup will roll forward
the store to the latest state (recorded in the logs).
Consistency check
A backup of the Microsoft Exchange Server database is considered as successful only if the
consistency check of the replicated datafiles succeeds. The consistency check is enabled by default.
To disable the consistency check, click on a created backup specification, right-click Microsoft
Exchange Writer in the Source tab, and then click Additional options. In this page, you can also
specify to throttle-back the consistency check for a second after the specified number of input/output
operations.
The consistency check can also be run before instant recovery.
Restore
The following scenarios are possible:
l
l
One or more databases are corrupted, but the log files are not damaged. In this case the database is
restored and transaction logs are applied—a rollforward recovery from the loss of one or more
databases.
The log files are corrupted or missing. In this case all databases and log files need to be restored. A
rollforward recovery of the database is not possible—a point-in-time restore after loss of a log file.
Rollforward recovery from the loss of one or more databases
For a rollforward recovery:
1. Dismount all stores from the storage group in which the target store resides using Microsoft
Exchange System Manager.
2. In the Data Protector GUI, switch to the Restore context. Expand Restore Objects and MS
Volume Shadow Copy Writers and select the client from which you want to restore the data.
In the Results Area, expand Microsoft Exchange Writer and select the stores you want to recover.
The Logs component is shaded and cannot be selected.
Note: Individual stores (databases) are always restored from the latest backup. The
Properties menu, which enables you to specify backup version, is not available at store level.
Selecting Microsoft Exchange Server stores for rollforward recovery
HPE Data Protector (9.08)
Page 70 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
3. Proceed as with general VSS writer restore. See "Standard restore" on page 44 for the general
VSS writer restore procedure.
4. Mount all stores from the storage group in which they reside using Exchange System Manager.
Selected stores are recovered.
Point-in-time restore after loss of a log file
To perform a point-in-time restore:
1. Start Exchange System Manager and check if the storage group is already dismounted. If not,
dismount the whole group.
2. Switch to the Restore context. Expand Restore Objects and Microsoft Volume Shadow Copy
Writers and select the client from which you want to restore the data.
In the Results Area, expand Microsoft Exchange Writer and select the whole storage group. Do
not select individual stores.
Selecting Microsoft Exchange Server stores for point-in-time restore
HPE Data Protector (9.08)
Page 71 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
3. Proceed as with general VSS writer restore. See "Standard restore" on page 44 for the general
VSS writer restore procedure.
4. Mount the stores from the storage group in which the target stores reside using Exchange System
Manager. All stores are mounted and put in the state as they were at the last selected full,
incremental, or differential backup.
Limitations
l
Shadow copies cannot be restored to alternate locations on the backup system.
l
You cannot restore the shadow copy to the Recovery Storage Group.
l
Rollforward recovery cannot be performed after a point-in-time restore.
l
Restore to Microsoft Exchange Server 2007 from backup made with Microsoft Exchange Server
2003 is not supported.
Instant recovery
With Microsoft Exchange Server 2003, you can restore a whole storage group or individual stores only
to the original location. With Microsoft Exchange Server 2007, you can restore a whole storage group or
individual stores to the original location or to a different location. For details, see "Microsoft Exchange
Server 2007 writer specifics" on page 76.
Note that restore to the original location will fail if not all objects residing on the target volume are
selected for restore. For example, if you have four stores and transaction logs on the target volume, and
HPE Data Protector (9.08)
Page 72 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
you select only one store for restore, the restore session will fail. Thus, the following configuration
scenarios are possible when restoring Microsoft Exchange Server data:
l
Transaction logs and database stores are on the same target volume.
You cannot select only database stores for instant recovery in the GUI or CLI. If transaction logs
and/or database stores are lost, the whole storage group (or all objects residing on the target volume)
need to be recovered.
In this case, you can perform only point-in-time recovery. Transaction logs will be replaced with the
backed up transaction logs.
l
Transaction logs and database stores are on different target volumes.
You can select only the database stores for instant recovery in the GUI or CLI. If a database store is
lost, it can be recovered separately provided that it resides alone on the target volume to be
recovered. Otherwise, all stores residing on the target volume must be selected for restore. If
transaction logs are lost, the whole storage group should be recovered.
In this case, you can perform either point-in-time or rollforward recovery.
To perform a point-in-time recovery of Microsoft Exchange Server writer data, select the whole
storage group. Transaction logs will be replaced with the backed up transaction logs.
To perform a rollforward recovery, select only the database stores and original location. The existing
transaction logs will be applied to restored databases. However, rollforward recovery will not be
possible if point-in-time recovery of the same backup session was performed before.
Prerequisites
l
Manually dismount the databases to be recovered.
Post-instant recovery steps
l
Manually re-mount the database stores.
Database recovery
You can run a database recovery from the Instant Recovery context of the Data Protector GUI. This
option is available, if you have created a separate backup specification for Incremental/Differential
backup with the same object and description as you have in the backup specification for instant
recovery. Such an Incremental/Differential backup is based on the Full backup with the selected
instant recovery option. You can select an Incremental/Differential backup in the Instant Recovery
context and start restore. Instant recovery will be performed and transaction logs will be automatically
applied to the recovered storage group.
User scenario for Microsoft Exchange Server 2003
backup and restore
This section provides examples of backup and restore policies for Microsoft Exchange Server 2003.
Three examples are provided, one for ZDB (transportable VSS snapshots) and two for different types of
restore.
HPE Data Protector (9.08)
Page 73 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Example - VSS transportable backup
This example describes a backup scenario for Microsoft Exchange Server 2003 using VSS
transportable backup together with a disk array of the HPE P6000 EVA Disk Array Family. The data is
to be backed up on tape once a day for the storage group containing critical mailboxes and every
second day for all other storage groups.
The storage groups should be backed up separately.
The following example setup is possible:
l
l
l
A Microsoft Exchange Server 2003 is running on the application system and is connected to a disk
array of the HPE P6000 EVA Disk Array Family. A tape drive is connected to the application system
for restore purposes.
A separate backup system is connected to a tape library.
Two storage groups are configured on the Microsoft Exchange Server 2003, each containing two
stores.
The first storage group, named Lists_Group, contains company-wide mailing lists. The storage
group contains two stores, Staff_Store for general mailing lists and Team_Store for team mailing
lists.
The second storage group, named Mailbox_Group, contains a larger number of individual mailboxes
for employees, which are given higher priority. This way the mailboxes are still available even if
mailing lists are down. The mailboxes are set up in two separate stores, one store for each
department, and named Sales_Store and Support_Store.
l
Every storage group has a separate backup specification to reduce the time needed for the shadow
copy creation. See "ZDB related problems" on page 114 for details of how to avoid issues with too
long copy creation.
HPE Data Protector (9.08)
Page 74 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Microsoft Exchange Server 2003 storage groups
For the first storage group (Lists_Group), create the following backup specification:
1. Select VSS transportable backup as the backup type. Select the application and backup system
and specify Snapshot (Differential) as the replica type.
2. Expand the Microsoft Exchange Server 2003 writer and select the first storage group (Lists_
Group) for backup.
3. Using the Data Protector scheduler, schedule backup sessions for a VSS transportable backup at
night. In the Recurring box, specify Full backup Weekly on Wednesdays and Fridays and
Incremental backup Weekly on other days. Set the backup protection to 2 days.
For the second storage group (Mailbox_Group), create the backup specification that can be used for
instant recovery:
1. Select VSS transportable backup as the backup type. Select the application and backup system
and set the Track the replica for instant recovery option.
2. Expand the Microsoft Exchange Server 2003 writer and select the second storage group (Mailbox_
Group) for backup.
3. Using the Data Protector scheduler you need to schedule backup sessions for a VSS
transportable backup every day. In the Recurring box, select Daily, set Recurring options to 1
day and set the backup protection to 4 days.
HPE Data Protector (9.08)
Page 75 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Example restore scenario for Microsoft Exchange Server 2003
In this example the Microsoft Exchange Server 2003 is configured as in "Example - VSS transportable
backup" on page 74 and the backup policy given in the example was implemented.
Example - restoring a single store
The second store (Support_Store) in the second storage group (Mailbox_Group), which contains
user mailboxes, is damaged, but the transaction logs and other stores are not damaged. Therefore,
a rollforward recovery will be performed only for this store.
To perform a rollforward recovery, proceed as follows:
1. Start the Microsoft Exchange Manager and dismount all stores in Mailbox_Group, but do not
dismount the stores from the first storage group.
2. Start Data Protector, go to the Instant Recovery context and expand the application system.
Select the last backup session of the backup specification with enabled instant recovery
options and expand Microsoft Exchange Server 2003 Writer. Select Support_Store under
Mailbox_Group for instant recovery.
Click Restore.
3. After the session is restored, start the Microsoft Exchange Manager. Mount all stores in the
second storage group.
Example - restoring a complete storage group after loss of transaction logs
In this example, the first store (Staff_Store) and transaction logs in the first storage group (Lists_
Group) are damaged but the second store (Mailbox_Group) is intact. Since the transaction logs are
damaged, a point-in-time restore of the whole storage group will be performed.
To restore the complete Lists_Group storage group, perform the following steps:
1. Start Exchange System Manager and check if the first storage group (Lists_Group) is already
dismounted. If not, dismount the whole group.
2. Start Data Protector. In the Restore context expand the application system. Select the last
backup session and then Microsoft Exchange Server 2003 Writer. Select the first storage
group (Lists_Group).
Start the restore.
3. Start Microsoft Exchange Manager and mount all stores in Lists_Group.
Microsoft Exchange Server 2007 writer specifics
Concepts
This section gives details of additional Microsoft Exchange Server 2007 features.
HPE Data Protector (9.08)
Page 76 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Continuous replication
Microsoft Exchange Server 2007 offers two models of replication for data protection that are supported
by Data Protector.
l
local continuous replication (LCR)
With LCR, you can create and maintain an exact copy (LCR copy) of databases in a storage group.
LCR copies are used in the event of data corruption, since you can switch the Exchange server to
use LCR copies in only a few seconds. If an LCR copy is used for backup and is located on a
different disk from the original data, the load on a production database is minimal.
l
cluster continuous replication (CCR)
CCR has the same characteristics as LCR. The only difference is that in the CCR environment,
databases and transaction logs are replicated to separate servers. Therefore, CCR copies can be
used for disaster recovery. You can perform VSS backups on the passive Exchange Server node
where the CCR copy is located and thus reduce the load on the active node.
The replicated storage groups are represented as a new Exchange Server writer instance, Exchange
Replication Service. They are backed up in the same way as original or production storage groups.
If using LCR or CCR with Standby Continuous Replication (SCR) configured, backups can only be
performed on the source side of the SCR. Microsoft does not support backups on the SCR target side.
For further information on SCR and supported SCR configurations, see the Microsoft website.
Restore to original or another location
With Microsoft Exchange Server 2007 writer, you can restore your data not only to the original location
(from which the backup was performed) but also to a different location. You can restore:
l
A whole storage group
l
A single store
In both cases the respective LCR or CCR copies can also be restored.
You can restore data to:
l
The original storage group
l
A different storage group
l
l
A non-Exchange location – With this restore method, after the restore is completed, the Recovery
Storage Group (RSG) can be created automatically.
A recovery server – This restore method restores data to different client and different storage group.
If you restore to a different storage group, you can access single mailboxes or individual e-mail
messages at a different location without changing the original storage group content. Furthermore, if the
whole server is destroyed, restoring to a different Exchange Server system (recovery server) will
minimize the time window during which your mailboxes will be unavailable.
Backup
Backup types
The Microsoft Exchange Server Writer supports the following Microsoft Exchange backup types:
HPE Data Protector (9.08)
Page 77 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
l
l
Full - backs up databases, transaction logs, and checkpoint files. The transaction logs are truncated.
Incremental - backs up the transaction logs to record changes since the last full or incremental
backup. The transaction logs are truncated.
This backup type is not available with the VSS hardware providers.
l
Differential - similar to incremental backup, but the transaction logs are not truncated.
This backup type is not available with the VSS hardware providers.
l
Copy - similar to full backup, with the difference that the transaction logs are not truncated. This
backup type is not intended for use in recovering failed systems.
Limitations
l
A combination of VSS snapshot backups and non-VSS backups (for example, stream incremental
backups) is not supported.
l
You can back up only the whole server or full storage groups. Single stores cannot be backed up.
l
Circular logging must be disabled.
l
l
With Exchange Server 2003, only one VSS backup session can be running on the same application
system at once. With Exchange Server 2007, only one VSS backup session backing up a specific
storage group can be running on the same application system at once. Consequently, if you start
such a backup session while another one is in progress, the latter waits until the first one finishes.
You must create a separate incremental backup specification for each backup specification that will
be used for instant recovery (has the Track the replica for instant recovery option enabled).
For example, you cannot create the backup specification BSpec1 for the storage group SGroup1
with snapshots as the backup type and the backup specification BSpec2 for the storage group
SGroup2 with snapclone as backup type and then use only one incremental backup specification
containing both storage groups.
Recommendations
l
With the HPE SMI-S P6000 EVA Array provider, an Exchange storage group should not reside on
more than four source volumes. However, the recommended configuration is that transaction logs
reside on one source volume and the stores on another source volume. This configuration enables
you to perform a rollforward recovery if only the stores are lost.
Create one backup specification for each storage group and schedule them in a row.
Rollforward recovery requirements
To ensure that a rollforward recovery from your backup is possible, consider the following:
Standard restore : Transaction logs must be backed up to enable the rollforward operation.
Instant recovery : Rollforward of a storage group can be performed with incremental or differential ZDB
to tape of the same objects as were backed up in instant recovery enabled sessions (ZDB to disk or
ZDB to disk + tape). If the logs are still available on disk, the restore of last full backup will roll forward
the store to the latest state (recorded in the logs).
Consistency check
A backup of the Microsoft Exchange Server database is considered as successful only if the
consistency check of the replicated datafiles succeeds. The consistency check is enabled by default.
To disable the consistency check, click on a created backup specification, right-click Microsoft
Exchange Writer in the Source tab, and then click Additional options. In this page, you can also
HPE Data Protector (9.08)
Page 78 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
specify to throttle-back the consistency check for a second after the specified number of input/output
operations.
The consistency check can also be run before instant recovery.
LCR and CCR environments
In LCR and CCR environments, the replicated storage groups are represented as a new instance of
Exchange Server writer, Exchange Replication Service. The replicated storage groups are backed up
in the same way as original (production) storage groups.
You can select any combination of storage groups for backup. However, you cannot select the original
and its replicated storage group in the same backup specification. See " Selecting a replicated
Microsoft Exchange Server 2007 storage group " below.
Selecting a replicated Microsoft Exchange Server 2007 storage group
Cluster support in a CCR environment
In a CCR environment, a cluster node from which you want a backup to be performed can be selected,
regardless of which instance (Information Store or Replication Service) resides on this node. If you
select the cluster node, Data Protector backs up any available instance on this node ignoring the
selection of the instance in the GUI even if, for example, you select the replicated storage group
(Exchange Replication Service) as backup object.
To specify the cluster node from which you want to perform a backup of any instance residing on this
node, right-click an Exchange writer and click Additional options. In the MS Exchange additional
HPE Data Protector (9.08)
Page 79 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
options dialog box, select the node under Back up any available instance from node. See "
Additional options for Microsoft Exchange Server 2007 in a CCR environment " below.
Additional options for Microsoft Exchange Server 2007 in a CCR environment
When backing up a Replication Service instance, backup may fail due to any of the following reasons:
l
The selected node is not available.
l
The status of the storage group to be backed up is not "Healthy".
l
Data Protector is not running on the selected node.
l
Vssbar.exe cannot be started on the selected node.
To avoid the session failing, select the option Revert to active node on failure in the same dialog
box. The backup will be restarted on the original server (active cluster node) and the original storage
group will be backed up. This option is ignored during backup of an Information Store instance.
Prerequisites
l
Before creating a backup specification and before running a backup, ensure that the Exchange
Server 2007 copy status is “Healthy”. Otherwise, you cannot browse the objects to be backed up
during creating backup specification, or the backup will fail. Note that the status “Initializing” is not
acceptable.
Limitations
l
In CCR environments, if a replicated storage group (Replication Service instance) is selected for
backup, no other VSS writer (for example, filesystem writer or SQL Server writer) can be selected to
be backed up in the same backup session. Because in this case, two writers would be located on
different systems, but Data Protector limits VSS backup sessions to involve only one system.
Considerations
l
l
Transaction log files are truncated after each full or incremental Microsoft Exchange Server backup.
The LCR and CCR clustering technologies, however, guarantee that logs that have not been
replicated are not deleted. Thus, running backups in a mode that truncates logs may not actually free
space. This may happen if replication of logs has not completed yet.
Before using P9000 XP Array VSS hardware provider in the resync mode, consider all applicable
limitations of this mode. The limitations are listed in "HPE P9000 XP Disk Array Family" on page 31
and "Limitations" on page 89. Due to these limitations, it is recommended to rather use P9000 XP
Array provider in the VSS compliant mode, or P6000 EVA Array VSS provider, or VSS software
provider.
HPE Data Protector (9.08)
Page 80 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Restore
You can restore Microsoft Exchange Server data by performing a standard restore or instant recovery
session:
l
See "Standard restore" below.
l
See "Instant recovery" on page 88.
Limitations
l
Individual databases (stores) can only be restored from the latest backup. To restore the databases
to an earlier point in time, the complete storage group must be restored.
Standard restore
The following scenarios are possible:
l
l
l
One or more databases are corrupted, but the log files are not damaged. In this case the database is
restored and transaction logs are applied—a rollforward recovery from the loss of one or more
databases.
The log files are corrupted or missing. In this case all databases and log files need to be restored. A
rollforward recovery of the database is not possible—a point-in-time restore after loss of a log file.
Some data in a mailbox has been lost. For example, an e-mail has been deleted by mistake. Restore
of the mailbox to an earlier point in time is needed. For details, see "Restoring individual mailboxes"
on page 83.
Consistency check
Optionally, to specify options for the consistency check of a Microsoft Exchange writer, right-click the
writer and click Additional options.
Rollforward recovery from the loss of one or more databases
For a rollforward recovery:
1. Dismount all stores from the storage group in which the target store resides using Microsoft
Exchange System Manager.
2. In the Data Protector GUI, switch to the Restore context. Expand Restore Objects and MS
Volume Shadow Copy Writers and select the client from which you want to restore the data.
In the Results Area, expand Microsoft Exchange Writer and select the stores you want to recover.
The Logs component is shaded and cannot be selected.
Note: Individual stores (databases) are always restored from the latest backup. The
Properties menu, which enables you to specify backup version, is not available at store level.
Selecting Microsoft Exchange Server stores for rollforward recovery
HPE Data Protector (9.08)
Page 81 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
3. Proceed as with general VSS writer restore. See "Standard restore" on page 44 for the general
VSS writer restore procedure.
4. Mount all stores from the storage group in which they reside using Exchange System Manager.
Selected stores are recovered.
Point-in-time restore after loss of a log file
To perform a point-in-time restore:
1. Start Exchange System Manager and check if the storage group is already dismounted. If not,
dismount the whole group.
2. Switch to the Restore context. Expand Restore Objects and Microsoft Volume Shadow Copy
Writers and select the client from which you want to restore the data.
In the Results Area, expand Microsoft Exchange Writer and select the whole storage group. Do
not select individual stores.
Selecting Microsoft Exchange Server stores for point-in-time restore
HPE Data Protector (9.08)
Page 82 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
3. Proceed as with general VSS writer restore. See "Standard restore" on page 44 for the general
VSS writer restore procedure.
4. Mount the stores from the storage group in which the target stores reside using Exchange System
Manager. All stores are mounted and put in the state as they were at the last selected full,
incremental, or differential backup.
Restoring individual mailboxes
To restore an individual mailbox:
1. Open the Data Protector GUI. In the Context List, click Restore. Expand Restore Objects and
MS Volume Shadow Copy Writers and select the client from which you want to restore the
data.
2. In the Results Area, expand the Microsoft Exchange Writer and select the storage group that
contains this particular mailbox. Right-click the storage group, click Properties, and specify the
desired point in time.
Right-click the storage group and click Restore as. In the MS Exchange additional options dialog
box, select the option Restore to a non-Exchange location and create RSG. In the Original
drop-down list, select the database that contains the mailbox, click Add, and then click OK.
3. Proceed as described in the general VSS writer restore procedure.
4. After the database has been restored, mount the database: open the Exchange Management Shell
or some other GUI tool like the Exchange Server Disaster Recovery Analyzer Tool.
To list all available databases, execute:
[PS] C:\>get-mailboxdatabase
HPE Data Protector (9.08)
Page 83 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Name
---LCR_store1
sg3_store1
store1
sg3_store2
sg1_store5
sg4_store1
sg3_store2
Server
-----TPC181
TPC181
exchclu3
TPC181
TPC181
TPC181
TPC181
StorageGroup
-----------LCR_sg1
sg3_local
sg1
sg3_local
First Storage Group
sg4
DP RSG
Recovery
-------False
False
False
False
False
False
True
In the example above, the database that has been restored to the Recovery Storage Group DP RSG
is named sg3_store2.
To mount the database, execute:
[PS] C:\>mount-database -identity "TPC181\DP RSG\sg3_store2"
5. Suppose the mailbox in question belongs to the person called John Doe. To extract the mailbox
items from the restored database and move the items to the John Doe's mailbox to the folder
RSG, execute:
[PS] C:\>restore-mailbox -RSGMailbox "John Doe" -RSGDatabase "TPC181\DP R
SG\sg3_store2" -identity "Doe" -TargetFolder RSG
Confirm
Are you sure you want to perform this action?
Recovering mailbox content from the mailbox 'John Doe' in the recovery
database 'TPC181\DP RSG\sg3_store2' into the mailbox for 'John Doe
(Doe@dp2.com)'. The operation can take a long time to complete.
[Y] Yes [A] Yes to All [N] No [L] No to All [S] Suspend [?] Help
(default is "Y"):Y
Identity
DistinguishedName
DisplayName
Alias
LegacyExchangeDN
:
:
:
:
:
PrimarySmtpAddress
SourceServer
SourceDatabase
SourceGlobalCatalog
SourceDomainController
TargetGlobalCatalog
TargetDomainController
TargetMailbox
TargetServer
TargetDatabase
MailboxSize
IsResourceMailbox
SIDUsedInMatch
SMTPProxies
:
:
:
:
:
:
:
:
:
:
:
:
:
:
HPE Data Protector (9.08)
dp2.com/Users/John Doe
CN=John Doe,CN=Users,DC=dp2,DC=com
John Doe
Doe
/o=DataProtector/ou=First Administrative Gro
up/cn=Recipients/cn=Doe
Doe@dp2.com
TPC181.dp2.com
TPC181\DP RSG\sg3_store2
TPC136
TPC136
dp2.com/Users/John Doe
TPC181.dp2.com
TPC181\sg3_local\sg3_store2
40553300B
False
Page 84 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
SourceManager
SourceDirectReports
SourcePublicDelegates
SourcePublicDelegatesBL
SourceAltRecipient
SourceAltRecipientBL
SourceDeliverAndRedirect
MatchedTargetNTAccountDN
IsMatchedNTAccountMailboxEnabled
MatchedContactsDNList
TargetNTAccountDNToCreate
TargetManager
TargetDirectReports
TargetPublicDelegates
TargetPublicDelegatesBL
TargetAltRecipient
TargetAltRecipientBL
TargetDeliverAndRedirect
Options
SourceForestCredential
TargetForestCredential
TargetFolder
PSTFilePath
RsgMailboxGuid
RsgMailboxLegacyExchangeDN
RsgMailboxDisplayName
RsgDatabaseGuid
StandardMessagesDeleted
AssociatedMessagesDeleted
DumpsterMessagesDeleted
MoveType
MoveStage
StartTime
EndTime
StatusCode
StatusMessage
ReportFile
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
: Default
:
:
: \RSG\Recovered Data - John Doe - 02.1
0.2008 16:49:59
:
: 0441be6c-46f6-4d8f-8562-ab615731ae89
: /O=DATAPROTECTOR/OU=FIRST ADMINISTRATIVE GRO
UP/CN=RECIPIENTS/CN=DOE
: John Doe
: deb5029b-4737-4fea-8c2d-ece24007e75d
: 0
: 0
: 0
: Restore
: Completed
: 02.10.2008 16:50:06
: 02.10.2008 16:50:12
: 0
: This mailbox in the recovery storage group d
atabase has been restored to the target user
mailbox.
: C:\Program Files\Microsoft\Exchange Server\L
ogging\MigrationLogs\restore-Mailbox20081002
-164958-8141342.xml
Restored mailbox items
HPE Data Protector (9.08)
Page 85 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Restoring a LCR or CCR copy to the original location
If you restore an LCR or CCR copy to the original location, the restore will be performed to the original
database (Exchange Information Store) and not to the database copy (Exchange Replication Service).
1. Right-click a storage group, a store, or logs and click Restore as.
2. In the MS Exchange additional options dialog box, select the target location for the components
you want to restore: target server, target storage group, and target stores. The following options
are available:
l
Restore to a different store
This option is selected by default.
Select this option to select the target stores for each store to be restored (original stores). First,
select the target system from the Target server name drop-down list, and then choose the
desired pairs of stores by selecting entries in the Original and Target drop-down lists. Note that
only stores cannot be restored, so logs are automatically selected in the restore session to
different location.
l
Restore to a non-Exchange location
Select this option to restore your data to a non-Exchange location. In this case, the restored
data will not be managed by Exchange Server and a Recovery Storage Group (RSG) will not be
created. You can manually create an RSG after the restore session completes. First, select the
target system from the Target server name drop-down list, and then choose the desired stores
using the Original drop-down list.
l
Restore to a non-Exchange location and create RSG
Select this option to restore your data to a non-Exchange location. After restore, Data Protector
will create a Recovery Storage Group called DP RSG on the target server. The selected stores
and logs will be restored to this recovery group. First, select the target system from the Target
HPE Data Protector (9.08)
Page 86 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
server name drop-down list, and then choose the desired stores using the Original drop-down
list.
By default, the storage group is restored in the C:\Omni directory. To select another location use
the Restore into location option. Click Browse and select the desired location.
The selected directory must be either empty or you can specify to create a new directory. If the
directory is not empty, the restore session fails.
Note that you can only specify the restore location for the storage group and not for a specific
store.
Restore to different location options (Exchange Server 2007 Writer)
HPE Data Protector (9.08)
Page 87 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Instant recovery
With Microsoft Exchange Server 2003, you can restore a whole storage group or individual stores only
to the original location. With Microsoft Exchange Server 2007, you can restore a whole storage group or
individual stores to the original location or to a different location. For details, see "Microsoft Exchange
Server 2007 writer specifics" on page 76.
HPE Data Protector (9.08)
Page 88 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Note that restore to the original location will fail if not all objects residing on the target volume are
selected for restore. For example, if you have four stores and transaction logs on the target volume, and
you select only one store for restore, the restore session will fail. Thus, the following configuration
scenarios are possible when restoring Microsoft Exchange Server data:
l
Transaction logs and database stores are on the same target volume.
You cannot select only database stores for instant recovery in the GUI or CLI. If transaction logs
and/or database stores are lost, the whole storage group (or all objects residing on the target volume)
need to be recovered.
In this case, you can perform only point-in-time recovery. Transaction logs will be replaced with the
backed up transaction logs.
l
Transaction logs and database stores are on different target volumes.
You can select only the database stores for instant recovery in the GUI or CLI. If a database store is
lost, it can be recovered separately provided that it resides alone on the target volume to be
recovered. Otherwise, all stores residing on the target volume must be selected for restore. If
transaction logs are lost, the whole storage group should be recovered.
In this case, you can perform either point-in-time or rollforward recovery.
To perform a point-in-time recovery of Microsoft Exchange Server writer data, select the whole
storage group. Transaction logs will be replaced with the backed up transaction logs.
To perform a rollforward recovery, select only the database stores and original location. The existing
transaction logs will be applied to restored databases. However, rollforward recovery will not be
possible if point-in-time recovery of the same backup session was performed before.
l
Some data in a mailbox has been lost. For example, an e-mail has been deleted by mistake. Restore
of the mailbox to an earlier point in time is needed. You need to perform a point-in-time recovery,
which is available in both configurations, when the transaction logs and databases reside on the
same storage volume and when they reside on different storage volumes. The procedure for
restoring an individual mailbox is similar as when performing standard restore (see "Restoring
individual mailboxes" on page 83). However, you also need to consider instant recovery-specific
information that is described in this section.
Prerequisites
For restore to a different location, Inet must run under a domain account which needs to be a member
of the following groups on the local system:
l
Administrators
l
Exchange Server Administrators
While configuring restore to a different storage group in the Data Protector GUI, only mounted stores
are displayed in the Target drop-down list. To enable all stores for target selection, mount the
dismounted stores and click Reload in the MS Exchange additional options.
During restore to a non-Exchange location and creation of Recovery Storage Group (RSG), Data
Protector deletes RSG that may already exists at the target location.
Limitations
l
In CCR environments where original database resides on a different P6000 EVA Array than the
database copy, restore of the database copy to the location where original database resides will fail,
HPE Data Protector (9.08)
Page 89 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
since restore to a different disk array is not supported. In this case, perform manual failover of
Exchange Server before the restore.
l
Restore to a different location can be only performed using Microsoft Virtual Disk Service.
Therefore, with P9000 XP Array VSS hardware provider, restore to a different location is possible
only after the backup session was run in the VSS compliant mode.
l
With P9000 XP Array VSS hardware provider, if a backup of an LCR or CCR storage group copy is
run in the resync mode, restore of data to the original location (original database or Exchange
Information Store) fails.
The reason for such a behavior is that source volumes, on which the database copy or Exchange
Replication Service resides, are in the suspended mirror relationship with the ZDB replica, which
should be restored to the original database.
Under such circumstances, the restore process first unpresents the source volumes from the
Exchange Server database copy, and presents them to the Exchange Server original database
afterwards. These actions result in shortage of source disks presented to the database copy. After
restore, you need to set up your LCR or CCR environment again.
To enable such restore process, set the omnirc option OB2VSS_FORCE_INSTANT_RECOVERY to 1,
perform restore, and after it, manually set up the LCR or CCR environment.
l
With Microsoft Exchange Server 2007 running in a CCR environment, whose database copy resides
on the backup system, there are particular situations where you need to consider additional steps for
performing instant recovery:
l
l
l
When P9000 XP Array VSS hardware provider is used, and the application system, where the
production database resides, and the backup system, where the database copy resides, are
connected to different disks arrays that are not configured in the same SAN. In such a
configuration, the application system does not see LUNs on the backup system and the other
way round.
When P6000 EVA Array VSS hardware provider is used and the disk arrays of the application
system and the backup system are not controlled by the same Command View.
When the P9000 XP Array VSS hardware provider is used in the resync mode, regardless of the
number of disk arrays are used.
In the above situations, there is only one instant recovery scenario. For a successful instant
recovery, follow the steps:
a. Fail over Exchange Server. This will cause your production database to reside on the former
backup system, where the database copy resided before failover.
b. Perform instant recovery.
c. After the instant recovery session completes, fail back Exchange Server.
This action makes all backup LUNs available on the same backup system where the database
copy resides.
Note: It is recommended that both production and replication server systems are using disks of
the same disk array.
For additional restore limitations, which apply also to instant recovery, see "Limitations" on the
previous page.
HPE Data Protector (9.08)
Page 90 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Restoring an LCR or CCR copy to the original location
See "Restoring a LCR or CCR copy to the original location" on page 86.
Post–instant recovery steps
1. Manually re-mount the database stores. In case of restore to Recovery Storage Group, re-mount
the database stores in this RSG.
2. If you restored an LCR or CCR copy to the original database, perform additional steps:
l
In case of an LCR restore, it is recommended to seed the restored database to synchronize the
original database with its copy.
For more information, see the webpage http://technet.microsoft.com/enus/library/aa995973.aspx.
l
In case of a CCR restore, you may need to perform additional steps.
Caution: Do not move the clustered mailbox server to the other node (using the MoveClusteredMailboxServer cmdlet) without first completing the procedure below. If you
move the server at this point, data loss may occur.
Perform the following steps to enable normal operation of the original and copy databases:
i. On the application system, mount the restored store.
ii. On the passive node, where the copy of the database exists, delete transaction logs.
iii. On the passive node, use the Update-StorageGroupCopy cmdlet to seed the storage
group copy or to re-synchronize the original storage group and its copy.
iv. On the passive node, use the Resume-StorageGroupCopy cmdlet to resume the storage
group copy.
Database recovery
You can run a database recovery from the Instant Recovery context of the Data Protector GUI. This
option is available, if you have created a separate backup specification for Incremental/Differential
backup with the same object and description as you have in the backup specification for instant
recovery. Such an Incremental/Differential backup is based on the Full backup with the selected
instant recovery option. You can select an Incremental/Differential backup in the Instant Recovery
context and start restore. Instant recovery will be performed and transaction logs will be automatically
applied to the recovered storage group.
Troubleshooting
For general VSS troubleshooting, see "Troubleshooting" on page 109.
Problem
Backup session waits 10 minutes to finish
By default, Data Protector waits 600 seconds for Microsoft Exchange Server 2007 writer to stabilize.
Action
HPE Data Protector (9.08)
Page 91 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Although this stabilization delay is recommended by Microsoft, you can change the waiting period by
setting the OB2VSS_EXCHANGE_WRITER_STABILIZATION omnirc option. Specify the waiting period in
seconds.
For details on how to set the option, see the HPE Data Protector Help index: “omnirc options”.
Microsoft Exchange Server 2010 writer specifics
Introduction
This section describes Microsoft Exchange Server 2010 writer specifics, when the Data Protector
Microsoft Shadow Copy Service integration is used.
Note: It is recommended that you use the Data Protector Microsoft Exchange Server 2010
integration instead of the generic VSS integration.
The Data Protector Microsoft Exchange Server 2010 integration offers additional functionality and
simplifies backup configuration, enabling you to back up multiple database copies (active and
passive) from different system in DAG in one session, and so on. See the HPE Data Protector
Integration Guide and the HPE Data Protector Zero Downtime Backup Integration Guide.
Microsoft Exchange Server 2010 concepts
Microsoft Exchange Server 2010 introduces the Database Availability Group (DAG), which is an
evolution of the Exchange Server 2007 SCR and CCR concept. Each DAG can consist of up to 16
systems, which can host multiple active and/or passive database copies. Passive copies are kept
consistent with the active copy and the DAG heals itself automatically if data on a particular node gets
corrupted. For example, if a passive copy gets corrupted, the passive copy is re-seeded. If an active
copy can no longer be healed, one of the passive copies becomes the active copy.
Both, the active copy and the related passive copies represent the same database and can be
exchanged during restore. This means that you can restore an active copy from a passive copy backup
or the other way round.
The most common scenarios involving Data Protector would therefore include (but are not limited to):
l
l
l
backing up passive copies thus avoiding additional loads on the active copy during backup
restoring a lost or corrupt passive copy, avoiding large volumes of network traffic when passive
copies are re-seeded
restoring the database to a point in time (for example for investigation purposes)
See "Restore scenarios" on page 94 for additional restore scenarios.
Integrating Data Protector Microsoft Volume Shadow Copy
Service integration with Exchange Server 2010
Using this integration, you can back up the Microsoft Exchange Server 2010 writers. Two writers are
used for Exchange Server 2010 backup and restore:
HPE Data Protector (9.08)
Page 92 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
l
Microsoft Exchange Writer (used for active copies)
l
Microsoft Exchange Replica Writer (used for passive copies)
Using the Exchange writers, only databases from a physical node (system) in a DAG can be backed
up. The DAG—which is a virtual entity and not recognized by Data Protector—cannot be backed up.
Configuration
Prerequisites
l
The Data Protector MS Volume Shadow Copy Service Integration component must be
installed on all DAG nodes (systems where Microsoft Exchange Server 2010 is installed) which you
want to back up.
Licensing
As you need to back up individual clients, you need one online-extension license for each system on
which the Data Protector MS Volume Shadow Copy Integration component is installed.
Configuring
To configure the integration for ZDB and ensure that disk resolving works, it is recommended that you
run the following command on the clients that you will back up from:
omnidbvss –resolve –apphost ClientName
where ClientName is the name of the Microsoft Exchange Server 2010 system, for example
server1.company.com. For details, see the omnidbvss reference page.
If the command is not run in advance, Data Protector will automatically run this command during the
first ZDB session which may slow down the backup session.
Backup
With the Data Protector Microsoft Volume Shadow Copy Service integration you can back up the
Exchange writer or the Exchange Replica Writer on a physical client or individual Mailbox Database
copies (either active or passive). You cannot select the database files or logs of a Mailbox Database
individually.
The following Microsoft Exchange Server 2010 backup types are supported:
l
Full
l
Incremental (only with the VSS software provider)
l
Differential (only with the VSS software provider)
l
Copy
HPE Data Protector (9.08)
Page 93 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Limitations
l
l
l
l
From the GUI, the additional options Perform consistency check and Throttle check for 1 second
cannot be set for the Microsoft Exchange Replica Writer. The Perform consistency check option is
enabled by default.
For ZDB, the Strict and Non-strict configuration check modes are not supported. Select Disabled in
the Configuration check mode drop-down list to disable the check. When this option is disabled,
Data Protector does not check if there are files on the source volume that were not selected for
backup or instant recovery.
You cannot combine a backup chain from full and incremental backups from different Microsoft
Exchange servers. You can also not combine a restore chain from passive and active copies.
You must create a separate incremental backup specification for each backup specification that will
be used for instant recovery (has the Track the replica for instant recovery option enabled).
For example, you cannot create the backup specification BSpec1 for the database
MailboxDatabase1 with snapshots as the backup type and the backup specification BSpec2 for the
storage group MailboxDatabase2 with snapclone as backup type and then use only one incremental
backup specification containing both databases.
Creating a backup specification
To create a backup specification, follow the procedure described in "Creating backup specifications" on
page 33.
When selecting the backup object (database), consider the following:
l
l
To create a backup specification for an active copy, select the Microsoft Exchange Writer.
To create a backup specification for a passive copy, select the Microsoft Exchange Replica
Writer.
After a failover, a passive copy becomes an active copy and an active copy becomes passive. In such
a case, the database backup object will fail, as the backup specification is not automatically updated
and Data Protector tries to back up an active copy using the Exchange Replica writer which is intended
for passive copies. You must manually update the backup specification and select the correct writer.
Restore
Follow the standard restore procedure as described in "Restore" on page 44. The restore scenarios in
this section describe only Microsoft Exchange Server 2010 specifics.
Restore scenarios
Restoring a passive copy
You can restore a passive copy from a chain of active or passive copy backups.
Restore when both the database files and logs are lost
HPE Data Protector (9.08)
Page 94 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
1. Suspend the database replication using the Exchange Management Console.
2. Right-click the database that you want to restore and open the Properties window. In the
Maintenance page, select This database can be overwritten by a restore.
Enabling restore for a database
3. In the Data Protector GUI, select the Restore context. In the Scoping Pane click MS Volume
Shadow Copy and select the client from which the database was backed up. You can select any
client, not only the one to which you restore.
4. Depending on whether you backed up an active or passive copy, expand Microsoft Exchange
Writer or Microsoft Exchange Replica Writer and select the backed up database copy.
Selecting a whole database
To perform a restore to a different client than the one from which the backup was performed, go to
the Options page and select the client from the Restore to another client drop-down list.
Restore to another client
HPE Data Protector (9.08)
Page 95 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
5. Specify the device and media options. For details, press F1.
6. Start the restore.
7. Resume the replication.
Restore when only database files are lost
If only database files are lost, but the logs are still available on the system, you can restore only the
database files:
1. Suspend the database replication using the Exchange Management Console.
2. Right-click the database that you want to restore and open the Properties window. In the
Maintenance tab, select This database can be overwritten by a restore.
3. In the Data Protector GUI, select the Restore context. In the scoping pane, expand MS Volume
Shadow Copy and select the client from which the database was backed up. You can select any
client, not only the one to which you restore.
4. Depending on whether you backed up an active or passive copy, expand Microsoft Exchange
Writer or Microsoft Exchange Replica Writer and expand the database backup copy.
Select File. See " Selecting the File component " below.
Selecting the File component
To perform a restore to a different client than the one from which the backup was performed, go to
the Options page and select the destination client from the Restore to another client drop-down
list. See " Restore to another client " on the previous page.
HPE Data Protector (9.08)
Page 96 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
5. Specify the device and media options. For details, press F1.
6. Start the restore.
7. Resume the replication
Restoring an active copy
You can restore an active copy from a chain of full and incremental backups of an active or any passive
copy.
As the passive copy is not always immediately updated, a point-in-time recovery (database and
transaction logs files are restored) using such a backup image may not result in the last state of the
active copy as it was at the time of the backup.
Restore when both the database files and logs are lost
1. Dismount the database using the Exchange Management Console.
2. Suspend the database replication. In the Exchange management console, right-click the database
that you will restore and open the Properties window. In the Maintenance page, select This
database can be overwritten by a restore. See " Enabling restore for a database " on page 95
3. In the Data Protector GUI, select the Restore context. In the scoping pane, expand MS Volume
Shadow Copy and select the client from which the database was backed up. You can select any
client, not only the one to which you restore.
4. Depending on whether you backed up an active or passive copy, expand the Microsoft Exchange
Writer or Microsoft Exchange Replica Writer and select the database.
If the system you backed up from currently hosts the active copy, the backed up copy will be
automatically restored as an active copy.
If the system you backed up from currently hosts a passive copy (for example, if a failover
occurred), select the client which hosts the current active copy from the Restore to another
client drop-down list. See " Restore to another client " on page 95.
5. Specify the device and media options. For details, press F1.
6. Start the restore.
7. Mount the database using the Exchange Management Console.
8. Resume the replication.
Restore when only database files are lost
1. Dismount the database using the Exchange Management Console.
2. Suspend the database replication. In the Exchange management console, right-click the database
that you will restore and open the Properties window. In the Maintenance page, select This
database can be overwritten by a restore. See " Enabling restore for a database " on page 95.
3. In the Data Protector GUI, select the Restore context. In the scoping pane, expand MS Volume
Shadow Copy and select the client from which the database was backed up. You can select any
client, not only the one to which you restore.
4. Depending on whether you backed up an active or passive copy, expand Microsoft Exchange
Writer or Microsoft Exchange Replica Writer and expand the backed up database.
5. Select File.
HPE Data Protector (9.08)
Page 97 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
If the system you backed up from currently hosts the active copy, the backed up copy will be
automatically restored as an active copy.
If the system you backed up from currently hosts a passive copy (for example, if a failover
occurred), select the client which hosts the current active copy from the Restore to another
client drop-down list. See " Restore to another client " on page 95.
6. Specify the device and media options. For details, press F1.
7. Start the restore.
8. Resume the replication.
Point-in-time restore
For a point-in-time restore, both active and the passive copies must be restored to the same point in
time or you must manually perform a full re-seed for passive copies. To avoid a full re-seed of passive
copies, you must restore the active and passive copies from the same backup copy.
1. Dismount the database using the Exchange Management Console.
2. Right-click the database that you want to restore and open the Properties window. In the
Maintenance page, select This database can be overwritten by a restore. See " Enabling
restore for a database " on page 95
3. Suspend the database replication.
4. Restore the active copy. See "Restoring an active copy" on the previous page.
Then select the database backup copy to be used for restore, right-click the database and select
Properties. Select the point in time to which to restore the active copy.
Selecting a point in time
5. If you do not want to perform a full re-seed of the passive copies, restore the passive copies. See
"Restoring a passive copy" on page 94.
HPE Data Protector (9.08)
Page 98 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
When selecting the database backup copy to be used for restore, select the same database
backup copy that was used for the active copy, right-click the database, and select Properties.
Select the same point in time as was selected for the active copy.
6. Mount the active database using the Exchange Management Console.
7. Resume the replication.
Instant recovery
Use the standard instant recovery procedure as described in "Backup" on page 27. For general
prerequisites and conditions (for example dismounting the Microsoft Exchange Server databases,
suspending the replication), see "Restore scenarios" on page 94. Additional instant recovery limitations
are listed in this section.
Limitations
l
l
l
You can perform instant recovery only on the client where the ZDB was created. The only exception
is when performing an instant recovery of a P6000 EVA Array snapclone, which can be restored to a
different client using the Data Protector CLI.
Instant recovery of a P6000 EVA Array snapclone to a different client is not supported through the
Data Protector GUI.
With the HPE P9000 XP Disk Array Family provider in the resync mode or with P6000 EVA storage
systems, each connected to its own system, you cannot perform a point-in-time recovery from a
single ZDB session. To be able to recover both an active copy and passive copies to the same point
in time, you need multiple ZDB sessions that were performed at the same time, one for the active
and one for each passive copy.
Instant recovery of a P6000 EVA Array snapclone to a different
client using the CLI
1. Use the omnidbvss command to get details about the session that you want to use for instant
recovery. For example:
omnidbvss -get session 2009/12/16-10
===================================================================
Session ID:
2009/12/16-10
Barlist Name:
Xdisk_XPcompliant_trans
Bar Hostname:
server1.company.com
Backup Type:
FULL
Instant Restore:
TRUE
Disk-Only:
FALSE
Component Name
===================================================================
[0]
/Microsoft Exchange Writer(Exchange Information Store)/
Microsoft Information Store/Xdisk_XPbox/Logs
HPE Data Protector (9.08)
Page 99 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
[1]
/Microsoft Exchange Writer(Exchange Information Store)/
Microsoft Information Store/Xdisk_XPbox/File
The objects that were backed up in this session are listed under Component Name.
See the omndbvss reference page in the HPE Data Protector Command Line Interface Reference
for details on the omnidbvss command.
2. Run the omnir command:
omnir -vss -instant_restore -barhost ClientName1 -session SessionID -tree
TreeName1 [-tree TreeName2...] -destination ClientName2 [VSS_INSTANT_RECOVERY_
OPTIONS]
See the omnir reference page in the HPE Data Protector Command Line Interface Reference for
details on the omnir command.
To recover the components from the above example, run:
omnir -vss -instant_restore -barhost server1.company.com -session 2009/12/16-10
-tree /Microsoft Exchange Writer(Exchange Information Store)/Microsoft
Information Store/Xdisk_XPbox/Logs –tree /Microsoft Exchange Writer(Exchange
Information Store)/Microsoft Information Store/Xdisk_XPbox/File -destination
server2.company.com -conf_check disabled
Troubleshooting
For general VSS troubleshooting, see "Troubleshooting" on page 109.
Problem
After a failover in a DAG occurs, Data Protector reports that some Microsoft Exchange writer
components could not be found:
[Major] From: OB2BAR_VSSBAR@server5.company.com "MSVSSW" Time: 12/7/2009 1:16:40 PM
Failed to find component that would match tree:
'/Microsoft Exchange Replica Writer(Exchange Replication Service)'.
Action
You can solve the issue in two different ways:
l
l
Update the backup specifications to use the correct writers.
Perform a failover so that the database returns to the original state (active or passive), the issue will
no longer be reported.
Microsoft Hyper-V writer specifics
Concepts
Microsoft Hyper-V writer is the successor to the Microsoft Virtual Server 2005 writer and is supported
on Microsoft Windows Server 2008. It is a VSS writer with similar set of functionality as Virtual Server.
HPE Data Protector (9.08)
Page 100 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
With both, it is possible to perform backups and restores of virtual machines. With Hyper-V, it is
possible to perform online backups using hardware providers.
For backup and restore specifics of Hyper-V writer, see "Backup" below and "Restore" on page 103.
Prerequisites
l
Windows Server 2008 Service Pack 2 must be installed. For more information, see
http://support.microsoft.com/kb/948465.
Backup
With the Hyper-V VSS writer, it is possible to back up:
l
The Hyper-V configuration
l
Virtual machines
Selecting Microsoft Hyper–V VSS writer backup objects
The following two types of backups are supported by the Data Protector VSS integration:
l
Online backup
Online backup of Hyper-V writer data is possible using a software provider or hardware provider.
If using a hardware provider for an online backup process, Hyper-V writer creates a shadow copy,
which is replicated together with the vhd file via the VSS hardware provider. Afterwards, the shadow
copy is presented to the Hypervisor System. In the case of transportable backup, the shadow copy
HPE Data Protector (9.08)
Page 101 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
is afterwards unpresented from the Hypervisor System and presented to the backup system,
specified in the backup specification.
After restore of an online backup session, the virtual machine is always turned off, regardless of the
state in which it was before the restore.
l
Offline backup
Offline backup is performed in the following cases:
l
The guest operating system – any other non-Microsoft guest operating system that is supported
by Hyper-V – is not VSS enabled.
l
The guest operating system does not have Hyper-V VSS integration services installed.
l
The virtual machine to be backed up is turned off.
Before offline backup, the virtual machine is automatically suspended (if not already) and resumed
after the backup.
After restore of an offline backup session, the virtual machine is suspended, regardless of the state
it was in before the restore.
The advantage of offline backup is that virtual machines are restored to the state they were in at
backup time, including the state of applications running at the time of backup. This is possible
because a virtual machine is in a suspended state during the backup.
Prerequisites
l
For online backup:
l
l
l
l
l
The guest operating system must have the Hyper-V VSS integration services installed and
should not use dynamics disks.
The snapshot file (avhd file) needs to be configured on the same volume as the virtual disk file
(vhd file).
The virtual machine to be backed up must be online.
Automatic mounting of new volumes must be enabled on the host (hypervisor) system. To enable
automatic mounting, execute MOUNTVOL /E on the hypervisor system.
For offline backup, if the prerequisites for online backup are met, a virtual machine to be backed up
must be put in the offline or suspended state manually.
Limitations
l
Cluster-aware backups are not supported.
l
ZDB to disk, ZDB to disk+tape, and instant recovery of a cluster node is not supported.
l
l
Only whole virtual machines can be backed up or restored. However, with ZDB to disk and ZDB to
disk + tape only a whole source volume can be backed up, regardless of the number of virtual
machines residing on the volume.
Backup of virtual machines configured to use physical disks is not supported.
HPE Data Protector (9.08)
Page 102 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
l
l
Backup of dismounted disks is not supported. A possible workaround is to configure mount points
for these volumes as described in: support.microsoft.com/kb/947021.
ZDB to disk, ZDB to disk+tape, and instant recovery of online virtual machines are not supported for
HPE P4000 SAN Solutions.
Backup from a physical cluster node
When backing up from a cluster node consider the following:
l
Offline backups trigger a failover.
During offline backups the virtual machine to be backed up is suspended for a moment. The cluster
server recognizes this as a failure and initiates the cluster failover. To avoid such a failover, perform
one of the following:
l
l
l
Run only online backups.
Before running an offline backup, manually put the virtual machine into the “Saved” state using
Failover Cluster Administrator.
Whenever a failover happens, you need to use another backup specification.
After a failover, the hostname where the virtual machine is running changes. Since this change is not
reflected in the original backup specification, you need to create a new backup specification to
specify the new hostname as the application system name.
Restore
Restore of Hyper-V writer data is possible to the original or to a different location. During a restore to a
different location, the Hyper-V writer checks whether a virtual machine with the same identity already
exists on the system. If it does, the Hyper-V writer removes the virtual machine from the system before
restore and imports the restored virtual machine. If such a virtual machine does not exist on the
system, it means that a restore session to this system is in progress or that the virtual machine was
already removed. It is possible to perform restore or instant recovery of Hyper-V virtual machines to
any Hyper-V system, which has a Hyper-V writer.
Restore from a physical cluster node
ZDB to tape sessions from a physical cluster node can be restored to any cluster node using a standard
restore procedure. To successfully restore such a session to a cluster node, perform the following
steps:
Backup sessions from a physical cluster node can be restored to any cluster node. To successfully
restore such a session to a cluster node, perform the following steps:
1. Delete the cluster group of the virtual machine to be restored using Failover Cluster Management.
2. Perform standard restore of the virtual machine on the cluster node where the virtual machine disk
is active. For the standard restore procedure, see "Standard restore" on page 44.
3. Re-create the virtual machine cluster group using Failover Cluster Management.
HPE Data Protector (9.08)
Page 103 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
Troubleshooting
Problem
Backup session of a Microsoft Hyper-V virtual machine ends unexpectedly
When backing up a Microsoft Hyper-V virtual machine, the session ends unexpectedly with an error
similar to the following:
[Major] From: OB2BAR_VSSBAR@computer.company.com "MSVSSW"
Time: 2/1/2011 11:29:03 AM [145:575]
Writer 'Microsoft Hyper-V VSS Writer' failed to prepare files
for backup:
Reported state:
VSS_WS_FAILED_AT_POST_SNAPSHOT
Expected state:
VSS_WS_WAITING_FOR_BACKUP_COMPLETE
Failure code:
VSS_E_WRITERERROR_NONRETRYABLE
The following are possible causes:
l
l
Automatic mounting is disabled on the host (hypervisor) system. For details, see
http://support.microsoft.com/kb/2004712.
There are issues inside the virtual machine, such as not enough free disk space for shadow copies,
use of a non-NTFS filesystem, and so on.
Action
Check if automatic mounting is enabled on the host hypervisor system. For example:
diskpart.exe
Microsoft DiskPart version 6.1.7600
Copyright (C) 1999-2008 Microsoft Corporation.
On computer: TPC021
DISKPART> automount
Automatic mounting of new volumes enabled.
If automatic mounting is disabled, enable it by executing the following command:
MOUNTVOL /E
If the issue persists even though automatic mounting is enabled, check the application logs inside the
virtual machine to determine the cause.
Microsoft SharePoint Services writer specifics
Concepts
The Microsoft SharePoint Services writer is a reference writer that integrates with the Windows VSS
framework, allowing backup applications to back up and restore Microsoft SharePoint data. This writer
has dependencies on:
HPE Data Protector (9.08)
Page 104 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
l
l
Search writers:
l
OSearch VSS writer
l
SPSearch VSS writer
SQL writers:
l
MSDE writer for Microsoft SQL Server 2000
l
Microsoft SQL 2005 writer for Microsoft SQL Server 2005
With the SharePoint Services writer you can back up and restore:
l
the configuration database
l
the central administration content database
l
other content databases
l
shared services provider databases
l
search databases
l
index files.
Backup types
The SharePoint Services writer supports the following Microsoft Office SharePoint Server 2007 backup
type:
l
Full (for databases and index files)
Limitations
l
Multi-server SharePoint configurations (farms) are not supported.
Prerequisites
l
The SharePoint Services writer and SQL writer are not started by default. Ensure that both writers
are properly installed and registered. The SharePoint Services writer must be registered using the
SharePoint command line administration tool:
stsadm -o
registerwsswriter
Backup
Limitations
l
If you back up individual writer components, to ensure that data is synchronized, the following
combinations of items must be backed up in the same session:
l
The configuration database and central administration content database
l
Search databases and corresponding index files
This means that, for example, you should not back up just a search database. Instead, always
select also the corresponding index file.
l
Do not use the standalone (not part of the reference) OSearch and SPSearch writers for backing up
the index files. If you use them, the search index will have to be reindexed.
HPE Data Protector (9.08)
Page 105 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
See " Selecting Microsoft SharePoint Services writer and the corresponding search writers" below for
an example of how to select writer components.
Selecting Microsoft SharePoint Services writer and the corresponding search writers
Restore
To ensure that data is synchronized the following items must be restored in the same session:
l
Configuration database and central administration content database
l
Search databases and corresponding index files
This means that, for example, you should not restore just a search database. Instead, always select
also the corresponding index file.
See " Selecting Microsoft SharePoint Services writer and the corresponding search writers for restore"
on the next page.
Because the configuration database and the central administration content database contain systemspecific information, you can restore them only to an environment that you configure to be precisely the
same, including all software updates, server names, and number of servers.
Prerequisites
HPE Data Protector (9.08)
Page 106 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
l
l
Before performing a restore, stop the following services:
l
Windows SharePoint Services Administration
l
Windows SharePoint Services Search
l
Windows SharePoint Services Timer
l
Office SharePoint Server Search
If you restore the whole farm, you must shut down the Internet Information Server (IIS).
Limitations
l
The VSS restore mode Restore files to temporary location is not supported.
l
The VSS restore option Restore to another client is not supported.
Selecting Microsoft SharePoint Services writer and the corresponding search writers for restore
HPE Data Protector (9.08)
Page 107 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 6: Writer specifics
MSDE writer specifics
Restore
MSDE writer is used to back up and restore Microsoft SQL database.
Before restoring the SQL system databases (master, model, msdb and pub), you have to stop the SQL
service.
MSDE writer
When you expand the MSDE Writer item in the Results Area, all Microsoft SQL Server instances are
displayed. Each instance contains all databases it includes. System databases (master, model, msdb
and pub) are always listed there.
If system databases are restored, the whole internal database structure will be changed.
Note: Only point-in-time restore is possible. Rollforward restore is not supported.
User databases will be restored only if it is possible to overwrite the files. MSDE writer will take the
user databases offline before the restore, while SQL service will have to be stopped manually in order
to restore the system databases.
HPE Data Protector (9.08)
Page 108 of 119
Chapter 7: Troubleshooting
This chapter lists problems you might encounter when using the Data Protector Microsoft Volume Shadow
Copy Service integration.
For general Data Protector troubleshooting information, see the HPE Data Protector Troubleshooting Guide.
Before you begin
l
l
l
Ensure that the latest official Data Protector patches are installed. See the HPE Data Protector Help
index: “patches” for information of how to verify this.
See the HPE Data Protector Product Announcements, Software Notes, and References for general Data
Protector limitations, as well as recognized issues and workarounds.
See https://softwaresupport.hpe.com/ for an up-to-date list of supported versions, platforms, and other
information.
Checks and verifications
l
l
l
On the application and backup systems, examine system errors reported in the debug.log file.
With the P9000 XP Array integration, ensure that RAID Manager Library is correctly installed on both the
application and backup systems. In addition, check if the libsvrrm.dll file exists in the RMLIB_home
directory.
Names of all clients in a Data Protector cell must match DNS entries and the Data Protector cell_server
file entries. If the names do not match, a warning message is displayed. In this case, do not use the VSS
integration as it can behave unpredictably. First reset network settings on the client and then re-import the
client.
Problems
Problem
Microsoft Exchange Server 2003 writer backup failed
When you start a backup of a Microsoft Exchange Server 2003 Writer, the following error is displayed:
[MAJOR]: Writer ‘Microsoft Exchange Writer’ failed to prepare files
for backup.
The reason may be the failure of the previous backup and, due to a Microsoft Exchange Server 2003 issue,
the Exchange Server Writer could not perform a proper cleanup.
Action
Restart the Information Store.
For more information, see the webpage http://support.microsoft.com/kb/945424/en-us.
HPE Data Protector (9.08)
Page 109 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
Problem
Microsoft Exchange Server 2003 aborts the backup if the shadow copy creation takes over 20
seconds
If an Exchange Server 2003 Writer is being backed up, the session can fail with VSSBAR reporting:
Snapshot could not be created.
In the application event log on the application system, the following event is recorded:
Event Type: Error
Event Source: ESE Event
Category: (16)
Event ID: 2004
Information Store (4916) Shadow copy 3 time-out (20000 ms).
Action
The following can help to solve the problem:
l
l
Limit the number of users that are accessing the management system.
Reduce the number of volumes in a snapshot set. Create a backup specification dedicated to each
storage group instead of one specification for the whole server. See "Backup" on page 69.
Problem
Backup of a Microsoft Exchange Server 2007 CCR database copy fails
During a backup session of a database copy in a Microsoft Exchange Server CCR environment, Data
Protector reports a major error notifying that the backup session has failed.
This problem may occur in CCR environments, due to a Microsoft Exchange Server 2007 issue with
incorrect display of database copy states in Exchange Management Console. In such a case, a
database copy in a "Failed" state may be displayed as "Healthy".
Action
To make the Exchange Management Console show the real status of a database copy, perform either
of the following actions:
l
Update Microsoft Exchange Server 2007 with Service Pack 1.
l
Perform re-seeding procedure as follows:
a. On the passive node, suspend the replication by using Suspend-StorageGroupCopy cmdlet.
b. Delete all log files from the Logs directory of the database copy.
c. Seed the database copy or re-synchronize the original database and its copy by using the
Update-StorageGroupCopy cmdlet.
d. Resume the database copy by using the Resume-StorageGroupCopy cmdlet.
e. On the passive node, check the status of the Exchange Replication Service by using vssadmin
list writers command. If the status is not stable, restart Microsoft Exchange Replication
Service.
Problem
Exchange Replication Service writer instance in LCR environment is not displayed in the Data
Protector GUI
HPE Data Protector (9.08)
Page 110 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
In the Data Protector GUI, while creating a backup specification, the Microsoft Exchange Writer
(Exchange Replication Service) object is not displayed on the pane for selection of backup
objects.
This problem may occur in LCR environments, due to a Microsoft Exchange Server 2007 issue with
incorrect display of database copy states in Exchange Management Console. In such a case, a
database copy in a "Failed" state may be displayed as "Healthy".
Action
To make the Exchange Management Console show the real status of a database copy and to enable
selection of the backup object, restart Microsoft Exchange Replication Service.
Problem
After the restore of system writers was aborted, the Windows operating system is corrupted
when you restart it
If the restore of some system writers (for example, System Writer) is aborted for any reason (hardware
or software failure, manually aborted, and so on), the Windows operating system may be corrupted
after the restart (for example, the GUI or some system services cannot be started, and so on).
Action
Depending on the nature of the corruption, repair or re-install the operating system from the Windows
installation CD-ROM.
Problem
Data loss during a point-in-time restore of a Microsoft Exchange Server 2003/2007 database
Though a point-in-time restore session finishes successfully, some of the data is not restored, because
existing logs may interfere with the mount procedure.
Action
Manually delete the log files before running the point-in-time restore.
Problem
Creation of an RSG fails when an RSG for the same storage group already exists
When you start a standard restore or instant recovery of the Microsoft Exchange Server 2007 writer
with the option Restore to a non-Exchange location and create RSG, the following error is
displayed:
[Major]
Application specific function 'PostRestoreEndExt' failed with error:
'The mailbox database that you specified is already associated with
a recovery mailbox database.
This error appears if an RSG for a storage group or mailbox database you restore is already created on
some other system.
Action
Since only one RSG can exist for the same storage group and since Data Protector can only delete an
RSG that exists on the target restore location, you need to manually delete the RSG on the other
system and restart the restore.
Problem
HPE Data Protector (9.08)
Page 111 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
VSS writers end up in Failed state after backup
After performing a backup session, VSS application writers may consistently end up in Failed state.
For example, during a backup of the Microsoft SQL Server, using the SQL 2005 writer, the writer fails.
After restarting the writer, the writer is in good state. However, during the next backup session, the
writer again fails.
The issue has been noticed for the SQL 2005 writer but may apply to other writers as well.
Use the vssadmin list writer command to check the state of the writers. For example, when you
run the command before the backup session:
vssadmin list writers
vssadmin 1.1 - Volume Shadow Copy Service administrative command-line
(C) Copyright 2001-2005 Microsoft Corp.
Writer name: 'SqlServerWriter'
Writer Id: {a65faa63-5ea8-4ebc-9dbd-a0c4db26912a}
Writer Instance Id: {c249df3c-f9d5-4f4a-8797-03724a60771c}
State: [1] Stable
Last error: No error
After the backup session:
C:\Program Files>vssadmin list writers
vssadmin 1.1 - Volume Shadow Copy Service administrative command-line
(C) Copyright 2001-2005 Microsoft Corp.
Writer name: 'SqlServerWriter'
Writer Id: {a65faa63-5ea8-4ebc-9dbd-a0c4db26912a}
Writer Instance Id: {e3d115c5-9c1a-47f6-8404-bfdba4c37890}
State: [8] Failed
Last error: Non-retryable error
Action
The reasons for such behavior often include incorrectly configured or expired user accounts, expired
passwords, and other connectivity issues between the application and the Volume Shadow Copy
Service. Ensure that your accounts are set up properly.
For Microsoft SQL Server, see the KB article http://support.microsoft.com/kb/919023 for information of
how to resolve the Microsoft SQL Server and VSS connectivity issues.
Problem
Backup or restore fail when Remote BLOB Storage (RBS) with the FILESTREAM provider is
used for SQL Server 2008
A backup session ends unexpectedly with the following error message:
[Major] From: OB2BAR_VSSBAR@computer.company.com
<mailto:OB2BAR_VSSBAR@computer.company.com>
"MSVSSW" Time: 2/3/2011 3:42:06 PM
[145:575]
Writer 'SqlServerWriter' failed to prepare files for
backup:
Reported state:
VSS_WS_FAILED_AT_PREPARE_SNAPSHOT
HPE Data Protector (9.08)
Page 112 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
Expected state:
Failure code:
VSS_WS_WAITING_FOR_BACKUP_COMPLETE
VSS_E_WRITERERROR_NONRETRYABLE
[Major] From: OB2BAR_VSSBAR@computer.company.com
<mailto:OB2BAR_VSSBAR@computer.company.com>
"MSVSSW" Time: 2/3/2011 3:42:06 PM
Cannot perform backup of:
'/SqlServerWriter(SQL Server 2008:SQLWriter)/BELMAVM20/SHAREPOINT/
FileStreamDB', which contains data in:
C:\temp\FileStreamDB\FileStreamDB.mdf
C:\temp\FileStreamDB\FileStreamDB_log.ldf
C:\temp\FileStreamDB\FileStreamData\*
Similarly, a restore session ends unexpectedly with the following error message:
[Major] From: OB2BAR_VSSBAR@computer.company.com
<mailto:OB2BAR_VSSBAR@computer.company.com>
"MSVSSW" Time: 12/8/2010 3:23:16 PM
[145:298]
Writer 'SqlServerWriter(SQL Server 2008 R2:SQLWriter)'
failed to prepare files for restore:
Reported state:
VSS_WS_FAILED_AT_PRE_RESTORE
Expected state:
VSS_WS_STABLE
Failure code:
VSS_E_WRITERERROR_NONRETRYABLE
The issues appear if FILESTREAM access level is set to Disabled.
Action
Ensure that FILESTREAM access level is set to Full access enabled or Transact-SQL access
enabled.
For details of how to configure RBS and FILESTREAM, see the Microsoft SQL Server 2008
documentation.
Problem
Microsoft Exchange Server restore or instant recovery fails
A Microsoft Exchange Server restore or instant recovery session fails with a message similar to the
following:
[Major] From: OB2BAR_VSSBAR@tpc202.company.com "MSVSSW"
Time: 19.02.2011 21:02:37
Post Restore for backup '2011/02/19-1' failed.
This may happen if a recovery of a Microsoft Exchange Server database lasts longer than the timeout
for post-restore operations (default is two hours). When the timeout is reached, Data Protector aborts
the session.
Action
Increase the timeout using the OB2VSS_WAIT_TIMEOUT omnirc option and restart the session. For
details on how to set the option, see the HPE Data Protector Help index: “omnirc options”.
Problem
The VSS integration can use only 5 concurrent threads for backup or restore
HPE Data Protector (9.08)
Page 113 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
During a backup or restore session, the VSS integration always uses only 5 concurrent threads even if
both the device and the application system are capable of handling more threads at once. This
limitations appears regardless of device concurrency settings.
Action
Modify the limit by setting the omnirc option OB2VSS_MAX_CONCURRENT_WORKER_THREADS to a higher
number. The maximum number of concurrent threads is 64.
ZDB related problems
Problem
Backup or instant recovery aborts due to VDS problems
A backup or instant recovery session aborts with the following error:
Failed to load VDS service.
This error can appear as a result of abnormal termination of a VDS service.
Actions
1. Stop the Virtual Disk Service (VDS):
l
l
l
Check that Virtual Disk Service is started, stop it using the command net stop vds or via the
Control Panel.
If the above step does not help, stop the Virtual Disk Service by terminating the process
vds.exe using the Task Manager. VDS will be started automatically as needed.
Alternatively, you could also log in to the system again and check if the system requests your
confirmation to stop an abnormally terminated VDS. If it does, reconfigure the debugger to
launch automatically to avoid similar problems in the future. This can be done by setting the
HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug value Auto to 1.
2. Check your configuration.
Successful Data Protector backup or restore completion is influenced by a number of factors
imposed by non-Data Protector components, including Microsoft Exchange Server and VSS. One
of the factors is timing, which in turn is impacted by environmental conditions such as I/O loads,
storage device activity and concurrent Data Protector actions.
To inhibit potential issues, it is recommended that you balance Data Protector activity across the
available activity or maintenance window. If you encounter issues, review and possibly reduce
concurrent Data Protector activities such as concurrent backups or restores.
Problem
No HPE SMI-S P6000 EVA Array provider login entries are configured within SMISDB
Action
Add the login information for HPE SMI-S P6000 EVA Array provider by executing:
omnidbsmis -ompasswd -add ClientName [-ssl] [-port PortNumber] [-user Username] [passwd Password]
Problem
HPE Data Protector (9.08)
Page 114 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
Configuration of an HPE SMI-S P6000 EVA CIMOM failed
Action
Execute the following command to perform a health check of your environment, which may help
identify any potential problems that occurred during the configuration:
omnidbsmis -ompasswd -check [-host ClientName]
Problem
Volume shadow copies cannot be imported due to low space in registry and consequently
backup fails
On a Windows system, one of the following errors is reported during a ZDB session:
l
By Data Protector:
[Critical] It was not possible to import Volume Shadow Copies from
application host.
l
By the operating system:
Error reported by the Windows Registry
These symptoms indicate that Windows Registry contains too many entries and ran out of space.
Such circumstances result in Data Protector ZDB session failures. Although Data Protector does not
write anything to the Registry directly, the Registry contains entries for all volumes ever presented to
the system. These entries are written by the storage device driver.
Action
Clean the Windows Registry up.
Windows Server 2003, Windows Server 2008: In the Command Prompt window, execute the
command mountvol /R.
It is recommended to periodically perform this cleanup task to prevent the Registry from running out of
space and Data Protector ZDB sessions from failing. However, the task should be run under
supervision and should not be automated.
Problem
Data Protector reports that a volume was not removed
During a backup session, the following error is reported:
Volume 'StorageID', which is part of backup 'backupID', was not removed.
This error may appear if a replica on the disk array is removed, but its entry in the VSS database
(VSSDB) remains intact.
This may happen when a backup session fails without creating a replica, but the reference that was
created for it could not be removed due to network problems.
HPE Data Protector (9.08)
Page 115 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
Action
To remove the false entry from the VSSDB:
1. In the error messages, find the ID of the session which could not remove the target volume.
2. Execute the following command, where SessionID is the session whose target volume(s) could
not be removed:
omnidbvss -remove session SessionID -reference
Upon successful removal, the command should display the confirmation message Removing
references of session SessionID from VSSDB.
Problem
Instant recovery of the Microsoft Exchange Server Writer fails
This problem may occur if the Microsoft Exchange Writer is not in the stable state. Check this by
executing VSSadmin list writers from the command prompt.
Action
Bring the Exchange Server Writer to a stable state by restarting the Microsoft Exchange Information
Store.
Problem
With a P6000 EVA Array, a copy-back instant recovery session completes with warnings
When you start a copy-back instant recovery session on P6000 EVA Array, the session completes
with warnings similar to the following:
[Warning] From: SMISA@seven.e2008.company.com "SMISA"
Time: 5/13/2010 11:07:13 AM [236:8040] A target volume is located
in a different folder than the source volume.
Source storage volume folder: Windows/VSSQA/Pump
Target storage volume folder: Windows/VSSQA/Seven
This warning is displayed when you perform an instant recovery of a storage volume that is located in a
different Command View EVA folder than the replica storage volume. As a result, when the instant
recovery completes, the newly-restored storage volume is automatically moved to the same folder in
which the replica storage volume is located.
Action
Using the Command View EVA, move the newly-restored storage volume back to its original location
to avoid inconsistencies with the Data Protector ZDB database.
Problem
After restore, the system restart error is displayed
In some cases, if the HBAs are changed on the application system, the software driver from previous
HBAs can cause this error after restore.
Action
Remove older drivers of uninstalled HBAs from the application system.
Problem
After an SQLServer writer instant recovery is restarted, the database cannot be brought online
HPE Data Protector (9.08)
Page 116 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
If an instant recovery of the SQLServer writer is aborted or fails, you can restart the session. However,
the SQLServer writer may report an error, stating that the files cannot be prepared during the instant
recovery and the session completes with errors.
The issue can appear if the disk was dismounted before the session was aborted and the files are not
visible to the writer. As a result, an error is reported and the database cannot be brought online after
restore.
Action
To recover the database after such session:
1. Detach from the database.
2. Attach to the database.
Problem
Shadow copy creation fails with the VSS system provider
When using the VSS system provider for a hardware LUN disk, the shadow copy creation fails.
Action
Install the latest vendor drivers and their support packs, including the HBA adapter and MPIO driver.
Problem
During a backup session, the volumes are imported and then immediately deleted
When you perform a backup using a hardware provider, the volumes that are being backed up are
imported and then immediately deleted with the following error:
[Normal] From: OB2BAR_VSSBAR@comp.company.com "MSVSSW" Time: 9/4/2010 1:37:41 PM
Imported Volume Shadow Copy with the properties: ...
[Normal] From: OB2BAR_VSSBAR@comp.company.com "MSVSSW" Time: 9/4/2010 1:37:41 PM
Deleting Volume Shadow Copies and releasing the volumes.
[Critical] From: OB2BAR_VSSBAR@comp.company.com "MSVSSW" Time: 9/4/2010 1:37:51 PM
Backup failed.
Action
Ensure that the client security settings are correct. If your cell is secured, the following systems must
be listed in the allow_hosts file:
l
For a local or network backup, the application system
l
For a transportable backup, both, the application and the backup system
For details, see the HPE Data Protector Help index: “securing, client systems”.
Problem
Instant recovery is not possible without P9000 XP Array VDS Hardware Providers
When the P9000 XP Array VSS hardware provider is in VSS compliant mode, the backup session ends
abnormally with the following error message:
VDS Swap Instant Recovery is not possible without VDS Hardware providers. Disk
backup will be aborted.
Action
HPE Data Protector (9.08)
Page 117 of 119
Integration guide for Microsoft Volume Shadow Copy Service
Chapter 7: Troubleshooting
To be able to perform instant recovery, install the VDS hardware provider:
1. Install the VDS hardware providers on the application and backup system.
2. Resolve the application system (the source volumes):
omnidbvss -resolve -apphost AppSystem
3. Restart the session.
Alternatively, if the ZDB backup will not be used for instant recovery, you can set the omnirc option
OB2VSS_ALWAYS_ALLOW_DISK_BACKUP_WITHOUT_VDS to 1.
If you perform a backup without the P9000 XP Array VDS hardware provider, instant recovery is not
possible.
Problem
Similarly, an instant recovery session ends abnormally with the message:
To perform VDS Swap Instant Recovery, the source and target LUNS need to be
resolved with VDS Hardware Providers.
Action
1. Install the VDS hardware providers on the application and backup system.
2. Resolve the application system (the source volumes):
omnidbvss -resolve -apphost AppSystem
3. Resolve the backup system (the target volumes created in the backup session):
omnidbvss -resolve -session SessionID
4. Restart the session.
Problem
Zero downtime backup sessions fail after updating the 3PAR StoreServ Storage firmware
After updating the firmware on a storage system of the HPE 3PAR StoreServ Storage family and
invoking a zero downtime backup session on this system, the session fails with an error similar to the
following:
[Critical] From: SMISA@appsys.company.com "SMISA" Time: 06/15/2012 2:23:28 PM
Replicator aborted with exception "No CIMOM found for ID 2FF70002AC000B6C".
Action
The cause of the problem is outdated information about the source volumes in the VSS database after
the storage system's firmware update. To solve the problem, update the VSS database by executing
the following command for the problematic application system:
omnidbvss —resolve —apphost ApplicationSystem
HPE Data Protector (9.08)
Page 118 of 119
Send Documentation Feedback
If you have comments about this document, you can contact the documentation team by email. If an email
client is configured on this system, click the link above and an email window opens with the following
information in the subject line:
Feedback on Integration guide for Microsoft Volume Shadow Copy Service (Data Protector 9.08)
Just add your feedback to the email and click send.
If no email client is available, copy the information above to a new message in a web mail client, and send
your feedback to AutonomyTPFeedback@hpe.com.
We appreciate your feedback!
HPE Data Protector (9.08)
Page 119 of 119
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising