Failover refers to the mechanism or capability within redundant computer systems whereby, if one component of the system fails or goes offline, another component automatically takes over its processing duties. The main objective of failover functionality is to achieve high availability (HA). Redundancy in systems is generally implemented through either HA clusters—primarily focused on improving availability—or Load Balancing (LB) clusters designed mainly for enhancing performance through workload distribution. Failover is specifically a feature associated with HA clusters.
The failover functionality is particularly favored for systems like database servers, where consistent operation and high availability are essential. Additionally, firewall systems—critical to ensuring internal network security—usually incorporate an HA cluster design with failover capabilities, commonly known as "stateful failover." Stateful failover specifically refers to the capability of transferring the active firewall's current session information directly to its standby firewall. If the active firewall encounters an issue, this ensures uninterrupted security by handing over processing seamlessly to the standby unit. To achieve stateful failover, two redundant firewalls must be interconnected using failover cables.
The F5 BIG-IP platform provided by F5 also supports this stateful failover functionality. Failover cables can be either dedicated serial cables or standard LAN cables. When serial cables are used for failover, the configuration is referred to as "Serial Failover," whereas failover through LAN cables is called "Network Failover."