Backups for Azure Stack HCI

You can perform backups automatically based on the configuration for a hypervisor or VM group, or manually for a VM group or a specific VM.

The first backup of a VM is always a full backup. By default, all subsequent backups are incremental, capturing any changes to VM data since the last backup.

You can recover virtual machine data, even when the most recent backup was incremental.

Backups run based on the following options:

  • Initial full backup: When you use Guided setup to set up the Virtualization solution, use the Back up now option to perform a backup for the default VM group.

  • Scheduled incremental backups: The server backup plan that is assigned to a VM group includes a schedule for ongoing incremental backups. Those backups are performed automatically based on the schedule, without requiring any user action.

  • Manual backups: You can perform on-demand backups for a VM group or for a specific VM.

If backups cannot start immediately, the backup jobs are queued.

What Gets Backed Up

  • Virtual machines (powered on or powered off)

  • VHD and VHDX files

  • Snapshot files

  • Configuration files for the virtual machines

  • Metadata that is required for granular recovery of files (NTFS and ext3 volumes only)

  • VHDX formatted hard disks with a sector size of 4KB without metadata

  • Azure Stack HCI replica VMs that are stopped

  • Backups enabled with Changed Block Tracking (CBT)

What Does Not Get Backed Up

  • Pass-through disks

  • Hard links (only for disk-level backup)

  • Virtual machine smart paging files

  • Virtual machines that the Volume Shadow Copy Service (VSS) fails to create a shadow for

  • Devices residing in a guest virtual machine over iSCSI or vHBA

  • Azure Stack HCI replica VMs that are running

  • Page and swap files (Azure Stack HCI Integration Services must be installed on guest VMs)