Site Recovery Failover and Reprotection Cycle

Azure Site Recovery keeps applications running during outages by replicating workloads to a secondary site. Recovery Plans organize how multiple virtual machines fail over together for complex applications, automating the startup order (e.g., database before web server) and allowing for custom scripts or manual steps.

There are three main failover types. A Test Failover is a drill that doesn't impact production. A Planned Failover for maintenance ensures zero data loss. An Unplanned Failover responds to a sudden disaster, aiming for minimal data loss. When triggering a failover, you select a Recovery Point—such as the Latest for the lowest RPO, Latest Processed for a lower RTO, or Latest App-consistent for database integrity. After a successful failover and verification, you Commit the operation to finalize it. To return to the primary region, you perform Reprotection to start reverse replication, followed by a Failback.