Fault tolerant PostgreSQL cluster with automatic failover using Repmgr and Barman. Part 1. Introduction.
Right off the bat, this is what you’ll get in the end: Synhronous standby server using repmgr Repmgr daemons running not on db servers No need of witness server with own PostgreSQL instance Split-brain tolerant auto-failover High available configuration without additional utils, such as HAProxy Backup server using barman Asynchronous WAL streaming Became synchronous, if …