VM-Centric Operations for Virtual Server Agent with VMware (VSA V2)

The Virtual Server Agent (VSA) for VMware supports VM-centric backup and restore operations using Indexing Version 2. (This feature is also called VSA V2 or VSA Indexing V2.)

In Commvault Version 11, Service Pack 14, this support is available for streaming and IntelliSnap backup operations performed using the Virtual Server Agent (VSA) with VMware.

For a new Commvault deployment, Indexing Version 2 is enabled by default, and new virtualization clients automatically use Indexing Version 2 to support VM-centric operations. For an existing Commvault deployment, you must enable Indexing Version 2.


  • Collecting file and folder information during backups is not supported with Indexing Version 2. Browse and restore operations use the live browse method to get file and folder information.

  • To restore guest files and folders from a VM-centric backup, use the disk library for one of the copies (primary or secondary) and perform a live browse from the disk library. Guest file restores from tape are not supported with VM-centric backups, because file details are not collected in VM-centric backups, and live browse operations are not supported from tape copies. As a workaround if no disk library is available, you can restore a full VM or a virtual machine disk from a tape copy.

Key Features

  • VM clients:

    • Automatically create clients, backup sets, and subclients for each virtual machine during discovery. These entities are visible to end users as well as administrators.

    • Associate VM backup sets to the parent backup set.

    • Enable or disable activity (backup, restore, and data aging) for individual VMs.

    • Specify security settings for individual VMs. In the CommCell Browser, end users can only see clients for virtual machines for which they have appropriate ownership and permissions.

    • Delete a VM client to delete the backup history for that VM.

    • Deleting a VM from the vCenter affects data retention and backup copy operations for the VM. For more information, see Data Aging and Retention Operations.

  • Backups:

    • Perform streaming or IntelliSnap backup operations and backup copy operations.

    • A backup operation for a subclient automatically creates child jobs with unique IDs for each VM, and commits VM-specific backup data as each child job finishes.

    • Distribute VM backup jobs across the available VSA proxies and MediaAgents, with a separate stream for each job.

    • Index extents are backed up for each VM. Subclient indexes are archived after all VM backups finish.

    • Changed block tracking (CBT) information is included as part of the index metadata. Cyclic redundancy check (CRC) information is stored as part of the extent header for each extent, and changed extents are backed up.

    • Run synthetic full backup operations for individual VMs.

    • Perform multi-stream synthetic full backup operations based on schedule policy configuration.

    • Perform application-aware backups.

    • For streaming backups, incremental backups are maintained when VMs are moved to a different subclient within the same backup set.

  • Job Control:

    • Perform job control operations for each VM.

    • Set the job priority for each VM.

    • For Indexing V2 Clients with Blackout Window operation rules configured, VM jobs will be marked FAILED that are started within the time windows specified.

  • Restores:

    • Browse and restore VMs or files from the client or subclient for an individual VM, or from the backup set for a subclient.

    • For a restore job that includes multiple VMs, the restore operation can use up to 10 streams (one stream per VM).

    • Customized options for end user restores.

    • Perform live mount operations and live recovery operations for an individual VM, from streaming backups, IntelliSnap backups, or backup copies.

    • Perform live sync operations for VMs from streaming backups, or Live Sync Direct operations from IntelliSnap backups that use the Virtual Server Agent Snap engine.

      Live sync operations must be configured for a subclient. You cannot configure a live sync schedule for an individual VM.

  • Other operations:

    • Customize retention and data aging for each VM.

    • Identify a subset of VMs for secondary copy operations.

    • Generate reports for specific VMs.