GLOSSARY / Resilience & DR

Failover

The automatic or manual process of switching operations to a standby system when the primary system becomes unavailable.

west Back to Glossary

Definition

sync_alt

Failover is the process, automated or manual, of redirecting operations to a backup system or region when the primary one fails. How quickly and cleanly a system fails over is what actually determines whether a stated RTO is achievable in practice; a disaster recovery plan that has never tested its failover process is, in effect, untested.

In Practice

A multi-AZ database configured for automatic failover can recover from a single availability zone outage in minutes without human intervention; a manually-triggered failover to a separate region typically takes considerably longer and needs a documented runbook to execute reliably.