The failback operation (restoring a VM from replica) allows you to restore your original VM by using the replica VM. This operation can also be used if you decide to stop the replica VM after the Failover operation is started and save the changes to the original or new location.
There are two options to recover - to an original or to a new location. If the original VM is missing or a new location is chosen then the new virtual machine will be created using the options specified during setting up the Failback VM from Replica task.
If the original VM is available and is chosen to be recovered to then this virtual machine will be stopped to perform the recovery from replica VM. Before starting the recovery the quiescied snapshot of the target VM using VMware API is created. This technical snapshot is required to be able to rollback in case the failback operation fails. If the new virtual machine is created then the snapshot is not required.
In case replica VM is running, the Failback VM from Replica task is restoring the original VM without stopping the replica VM. Only when the failback is finished, the replica VM is stopped. Finally, the Failback VM from Replica task recovers the changes from replica VM to the original (new) VM that have been done during the time of the failback operation. It allows to minimize the downtime and to restore the VM state as much as close to its replica state.