Backing Up Cluster Shared Volumes in a Failover Cluster in Windows Server 2008 R2

Updated: August 24, 2011

Applies To: Microsoft Hyper-V Server 2008 R2, Windows Server 2008 R2

There are multiple methods for backing up information that is stored on Cluster Shared Volumes in a failover cluster running Windows Server 2008 R2 or Hyper-V Server 2008 R2. This topic describes the two most common methods, virtual-machine based backup and management operating-system based backup. Management operating-system based backup is sometimes called parent-partition based backup.

Important

Be sure to review carefully what your backup application backs up. Also, for management operating-system based backup, ask your backup application vendor about the compatibility of your backup application with Hyper-V and with Cluster Shared Volumes.
You may require administrative credentials when backing up a failover cluster.

In this topic

Summary of the two backup methods

Virtual-machine based backup

Management operating-system based backup

Summary of the two backup methods

The following table summarizes the similarities and differences between virtual-machine based backup and management operating-system based backup:

Characteristics of virtual-machine based backup Characteristics of management operating-system based backup

Runs within a virtual machine, in the same way that a backup application runs within a physical server. When there are multiple virtual machines that are managed centrally, each virtual machine can run a backup "agent" that is controlled from the central management server.

Runs through Hyper-V and must work compatibly with Cluster Shared Volumes. When there are multiple virtual machines configured within Hyper-V, you can back up all of them, or a selection of them, in one operation.

Usually backs up only the application data for the application that is running in the virtual machine.

In addition to backing up application data (which is stored in a VHD), backs up the configuration of virtual machines, snapshots associated with the virtual machines, and all VHDs used by those virtual machines.

A backup application that is compatible with Hyper-V and CSV can back up multiple VHD files regardless of how they are arranged on the Cluster Shared Volumes.

The person running the backup application or agent needs standard backup privileges.

The person running the backup application needs administrative privileges on all failover cluster nodes.

The backup application or agent works in essentially the same way as it would in a physical server. It does not need to be written specifically for a CSV environment.

The backup application must use the VSS Hyper-V writer and must be compatible with CSV. See your application vendor for details about the compatibility of your backup application.

Virtual-machine based backup

The following illustration shows a virtual-machine based backup of the application data for Virtual Machine 1, which is configured in a failover cluster that is using Cluster Shared Volumes. The backup application runs within a virtual machine in the same way that a backup application runs within a physical server.

Figure 1   Virtual-machine based backup

Note that the way that the VHD files are arranged in the preceding example (with all the data VHD files placed together on one Cluster Shared Volume) is one of many ways that VHD files can be arranged on the Cluster Shared Volumes in a cluster. For more information, see Recommendations for Using Cluster Shared Volumes in a Failover Cluster in Windows Server 2008 R2.

When there are multiple virtual machines being managed centrally, each virtual machine can run a backup "agent" (instead of running an individual backup application) that is controlled from the central management server.

Management operating-system based backup

The following illustration shows a management operating-system based backup, also known as parent-partition based backup. The backup captures all the information about multiple virtual machines that are configured in a failover cluster that is using Cluster Shared Volumes. The backup application runs through Hyper-V, which means that it must use the VSS Hyper-V writer. The backup application must also be compatible with Cluster Shared Volumes. The backup application backs up the virtual machines that are selected by the administrator, including all the VHD files for those virtual machines, in one operation. In the following illustration, VM1 and VM2 are being backed up:

Figure 2   Management operating-system based backup

Note that the way that the VHD files are arranged in the preceding example (with all the data VHD files placed together on one Cluster Shared Volume) is one of many ways that VHD files can be arranged on the Cluster Shared Volumes in a cluster. For more information, see Recommendations for Using Cluster Shared Volumes in a Failover Cluster in Windows Server 2008 R2.