failback
also: switchback, fallback
The process of restoring service to the original server or resource after it has been recovered, following a failover event. Failback reverses the failover by moving traffic and operations back from the backup system to the primary system.
Failback is the return phase in a high-availability system. When a primary server fails, a failover automatically redirects traffic to a standby backup server. Once the primary server is repaired and brought back online, failback transfers control and traffic back to it.
Failback can be automatic or manual depending on the system configuration. Automatic failback resumes operations on the primary immediately upon recovery, while manual failback requires an administrator to explicitly trigger the switchback to give time for testing and validation.
For example, in a database cluster: Server A fails, so traffic automatically fails over to Server B. When Server A is fixed and running again, failback redirects all connections back to Server A, restoring the original configuration.
Failback is commonly managed by tools like pacemaker, corosync, keepalived, and cloud load balancers. It is distinct from failover in that failover responds to failure, while failback responds to recovery.