After the primary region has been recovered from a failure it may still contain stale production VMs and an old AROVA. Stale production VMs must be deleted from the primary region before initiating failback. If any VM exists with the same name in the target zone of the primary region, AROVA will not allow the failback operation to proceed.

    • Stale production VMs must be removed from the primary region before initiating failback.
    • Stale instances of AROVA must be cleaned up before initiating failback.
      • After the primary region has been recovered, any old instance of AROVA should be deleted and a new AROVA deployed to it.
      • If a stale instance of AROVA exists in the primary region, the AROVA deployment script will fail and display a warning message.
        • The script can be restarted using the same arguments after the stale AROVA has been deleted.

See:

AROVA Failback

VM Failback