Fail back VMware VMs to on-premises site

This article describes how to fail back Azure VMs to an on-premises site, following failover of on-premises VMs to Azure with Azure Site Recovery. After failback to on-premises, you enable replication so that the on-premises VMs start replicating to Azure.

Before you start

  1. Learn about VMware failback.
  2. Make sure you've reviewed and completed the steps to prepare for failback, and that all the required components are deployed. Components include a process server in Azure, an on-premises master target server, and a VPN site-to-site connection (or Express Route private peering) for failback.
  3. Make sure you've completed the requirements for reprotection and failback, and that you've enabled reprotection of Azure VMs, so that they're replicating from Azure to the on-premises site. VMs must be in a replicated state is order to fail back.

Run a failover to fail back

  1. Make sure that Azure VMs are reprotected and replicating to the on-premises site.
    • A VM needs at least one recovery point in order to fail back.
    • If you fail back a recovery plan, then all machines in the plan should have at least one recovery point.
  2. In the vault > Replicated items, select the VM. Right-click the VM > Failover.
  3. In Confirm Failover, verify the failover direction (from Azure).
  4. Select the recovery point that you want to use for the failover.
    • We recommend that you use the Latest recovery point. The app-consistent point is behind the latest point in time, and causes some data loss.
    • Latest is a crash-consistent recovery point.
    • With Latest, a VM fails over to its latest available point in time. If you have a replication group for multi-VM consistency within a recovery plan, each VM in the group fails over to its independent latest point in time.
    • If you use an app-consistent recovery point, each VM fails back to its latest available point. If a recovery plan has a replication group, each group recovers to its common available recovery point.
  5. Failover begins. Site Recovery shuts down the Azure VMs.
  6. After failover completes, check everything's working as expected. Check that the Azure VMs are shut down.
  7. With everything verified, right-click the VM > Commit, to finish the failover process. Commit removes the failed-over Azure VM.


For Windows VMs, Site Recovery disables the VMware tools during failover. During failback of the Windows VM, the VMware tools are enable again. If you disable replication after failback from Azure to on-premises, the seed disks will still be present in Azure. Ensure that they have been deleted to not incur any storage costs.

Reprotect from on-premises to Azure

After committing the failback, the Azure VMs are deleted. The VM is back in the on-premises site, but it isn't protected. To start replicating VMs to Azure again,as follows:

  1. In the vault > Replicated items, select failed back VMs, and then select Re-Protect.
  2. Specify the process server that's used to send data back to Azure.
  3. Select OK to begin the reprotect job.


After an on-premises VM starts, it takes up to 15 minutes for the agent to register back to the configuration server. During this time, reprotect fails and returns an error message stating that the agent isn't installed. If this occurs, wait for a few minutes, and reprotect.

Next steps

After the reprotect job finishes, the on-premises VM is replicating to Azure. As needed, you can run another failover to Azure.