Challenge
This article explains the steps to perform a manual failback using reverse replication when a normal failback fails or is not possible because the replica was powered on outside of the Veeam console.
This process essentially turns replication on its head and uses it as a tool to synchronize the state of the replica back on to a production VM.
Note: If you powered the replica on manually to a different snapshot than the bottom most snapshot you will most likely need to recreate the replica.
Solution
Important Note
The following steps will document how to manually failback a replica when the normal failback operation fails or if the replica was manually powered on. These steps will replicate the current state of the replica on to an existing production VM. If there is not a production VM to “failback” on to, you can skip the mapping steps (Part 1:Steps 2 & 6) and a VM will be created in the destination specified within the manual failback replication job.
──────────────────────────────────────────────────────────
Part 1: Create the new replication job
──────────────────────────────────────────────────────────
1. Create a new replication job.
2. Check the "Low connection bandwidth" box on the first page:
Note: This step can be skipped if you are not failing back to a production VM.
3. Add the replica to the job as the source VM in the Virtual Machines tab:
Note: If no production VM exists to failback to, specify where you want the state of the replica to be duplicated to.
5. On the Job Settings tab, choose a repository for the meta data that is closest the VM that was added on the Virtual Machines tab.
6. On the Seeding tab, select the “Map replicas to existing VMs” checkbox.
a. Highlight the replica, click edit and then select the production VM to be “failed back” onto.
8. Skip the Schedule tab, and click Finish to create the job.
──────────────────────────────────────────────────────────
Part 2: Manual Failback Procedure
──────────────────────────────────────────────────────────
The following steps are intended to be performed while the replica is powered on until the step that instructs you to power it off. If the replica is already powered off you can skip to step 3.
1. Run the replication job once.
Note: This step may take a very long time.
2. Cleanly shutdown the replica.
3. Run the replication job.
Note: This is done so that the failback target and the replica are in the exact same state, since the replica is powered off it will not change.
4. Within the Veeam Backup & Replication console navigate to Backup & Replication>Replicas>Ready
5. Find the entry whose job name matches the job created in Part 1 of this KB.
6. Right-Click and Select Failover Now… and Next, Next, Finish
7. The production VM will now power on, please verify its functionality then continue.
8. Right-click the same entry as Step 6 (Now under Replicas>Active) and select Permanent Failover.
Note: The permanent failover will consolidate all snapshots on the replica-of-the-replica-now-production vm.
──────────────────────────────────────────────────────────
Part 3: Cleanup and Return to normal operation
──────────────────────────────────────────────────────────
Note: Re-add the production VM to the normal replication job and any backup jobs it was part of previously, as its Ref-ID may have changed. Within the normal replication job enable the seeding option then use the mapping feature under the seeding tab to map the new production server to the existing replica.
•Delete the manual failback replication job created in Part 1.
•If you manually powered on the replica, run the regular job to force Veeam to return it to a normal state.
•If you failed over the replica from within the Veeam console, you will need to undo failover prior to running the job.