Automatically switching to a different, redundant system upon failure or abnormal termination of the currently active system. Failover can be applied to a cluster of servers, to network or storage components or any other set of redundant devices that must provide high availability because down-time would be expensive or inconvenient. It may be implemented in hardware, software or a combination.