Servers in a 1+1 High Availability cluster pair communicate with each other using a heart-beat synchronization mechanism. Interruptions in the heart-beat from the active server trigger a failover to the standby server. The failure of a critical component process on the active server also triggers a failover to the standby server. The Primary and Backup servers are identical in functionality and configuration, resulting in a seamless failover.

The system synchronizes the state of all active sessions to the Backup server in real-time. State synchronization ensures the Backup server preserves the state of each active session without interruption to the user. Scenarios where the session state synchronization might not be fully synchronized, are handled by notifying the application of the failover. The failover notification provides the application the opportunity to run proper recovery steps for the given session state, for example, re-prompting the user for digit collection.

Both the Backup and the Primary servers can become active at the same time if the servers become network isolated from each other. When the servers reconnect, the servers exchange state information. The system uses the state information to select the server that becomes the active server and the server that becomes the standby server. The system selects the server that was the last server to process a new session as the active server. If the system did not process a new session then the server that was active the longest becomes active. In most cases, the server that was active for the longest period is the server that was active before the network isolation occurred.

When the High Availability state is locked, the system prevents failovers and service redundancy is unavailable. You set the High Availability state to locked only when the 1+1 High Availability cluster is recovering from a network isolation issue. The Locked state ensures that sessions are not lost from the server that processes the sessions during the network isolation recovery.

Under normal conditions, the High Availability state must not be unlocked. After the network isolation issue is resolved and both servers are actively part of the cluster, ensure that the High Availability state is not locked. Both servers must be unlocked to provide failover redundancy.