Hi everyone I run a PostgreSQL 14 environment with one primary and two standby instances with streaming replication. I recently had a case where the streaming replication on one of the standby instances paused (seemingly automatically):
2024-06-01 01:11:38.482 CEST [4575476] @ app=LOG: recovery has paused 2024-06-01 01:11:38.482 CEST [4575477] @ app=HINT: Execute pg_wal_replay_resume() to continue. The log messages around this time all look normal and there were no outages. Are there any circumstances under which replication would automatically pause without outside intervention? There was no change of the primary during this time (we do not load balance queries to the standby nodes, they exist soley for HA). I manually resumed using the suggested command and it's been running fine since (and it's been running for about a year until the pause without issue the other day). For reference, the full environment consists of the three PostgreSQL nodes, with EnterpriseDB Enterprise Failover Manager (but as far as I can tell this didn't intervene at the time either), and three PgPool nodes. Many thanks Ian van der Linde
