You can run a recovery plan for failover immediately after a real life disaster event, or run it as a test failover before a real disaster occurs.
You can run a recovery plan for failover in the following ways:
- Failover. A failover operation is run following a disaster event when the source site is no longer available. The failover operation orchestrates on the destination site based on previously replicated snapshots. When failing over to a recovery SDDC, VMs that belong to the protection groups defined in your recovery plan are recovered to the vCenter on the recovery SDDC. With a failover, you have the option to commit, cancel and rollback, or terminate. When a plan is committed, you can then create a failback plan to be used later during failback once the source site is recovered.
- Test failover. A test failover operation is similar to regular failover operation, but runs in the context of its own test execution environment only if the recovery plan has been configured to have separate configuration mappings for failovers and test failovers. Another difference is that by default, a test failover stops execution with every encountered error. In an actual failover, the default behavior is to continue running the plan even, if errors are found during recovery. You have the option to override the default behavior for both failover types by changing the default runtime settings prior to starting the failover operation. Finally, with a test failover your only option is to clean up the test plan. There is no failback function or capability to fail workloads back to the original protected site. For more information, see Test Recovery.