The failback process consists of three general stages: undo, catchup, and completion.
Power Off and Undo Stage
In the Undo stage, VMs on the failback target are restored to the state that matches snapshots used at recovery time.
The Undo stage of a failback ensures that all VMs protected by a recovery plan power off and their state matches the state of the snapshot selected for failover to arecovery SDDC. The Undo stage is necessary because VMs on the protected site might not be powered off prior to failover (for example, if the site became disconnected). These VMs continue to run and accumulate any arbitrary changes that diverge with the authoritative VM state in the recovery SDDC.
In the Catchup stage, VM changes that have occurred while running on a recovery SDDC following failover are applied to the VMs on the failback target site
The Catchup stage of a failback operation applies incremental changes of all modified VM state in the recovery SDDC to the VM copies on the failback target. The duration of this stage depends on the change rate and the amount of time VMs are running on the Recovery SDDC.
During this stage, VMs remain in powered off state on both the failback target and on the recovery SDDC. Because of service downtime during this stage, it might be desirable to schedule it during the maintenance window. For example, you can edit the failback plan and inserting a wait-for-user prompt step prior to running the catchup.