Postgres Failover and Replication Manager

Detect compute, network, and database failures quickly to automatically promote the most up to date standby and reconfigure other standbys to point to the new master.

Ensure Business Continuity

Minimize unplanned downtime for your applications and maintain high availability.

Maintain Service Level Agreements

Quickly detect failure by continuously monitoring health of physical replication nodes.

Peace of Mind

Avoid split-brain scenarios and time lost from triggering false negatives or unnecessary failovers.

Architecture

Agents run on the master and standby nodes continuously monitoring their health. When a failure is detected, agents send alerts, and initiate an automatic failover or wait for user to take the appropriate action. When there are only two nodes, a witness node is used to confirm assertions and break the tie in the event of a conflict.