Glossary · System coordination, integration and orchestration
Switchover
Also known as: Failover, Redundancy switchover
German: Umschaltung
In redundant systems, a switchover is the transfer of the active role from one component, such as a controller, server, network path or power supply, to its redundant counterpart, either automatically after a failure or manually for maintenance.
- System integration
In one sentence
A switchover transfers the active role from one redundant component to its counterpart, automatically after a failure or manually.
Example
When the primary controller of a redundant pair fails, the standby takes over within one cycle and the process continues without interruption.
How it applies
- Engineering: Specify switchover time, what state is preserved, which outputs may change during switchover and how the process tolerates it. Test that the process really tolerates this behavior.
- Commissioning: Test switchover under load, repeatedly and in both directions. Also test failure of the standby, which may go unnoticed without monitoring.
- Operation: After a switchover, the system runs without redundancy until the failed component is repaired. Treat this like a Degraded mode with alarms and a repair deadline.
- Documentation: Describe switchover behavior, alarms, manual switchover procedures for maintenance and the steps to restore redundancy.
Switchover vs. split brain
A switchover transfers the active role cleanly: the old primary stops acting as primary. A Split-brain condition happens when both components believe they are active, usually because the communication between them failed rather than one of the components itself.