During a high traffic portion of the day, one of your relational databases crashes, but the replica is never promoted to a master. You want to avoid this in the future. What should you do?

Routinely scheduled failovers validate your disaster recovery process and ensure that your replica can successfully be promoted to a master. This practice uncovers configuration issues or operational gaps before a real incident occurs. Using a different database or choosing larger instances doesn't address the underlying failover mechanism's reliability. Creating snapshots more regularly is good for data recovery but doesn't test the failover process itself.