06. Replication and HA

Monitor replication lag, slot activity, and the overall high availability posture of your PostgreSQL cluster.

note

This dashboard is currently under development. In the live demo, it shows a placeholder. The metrics are being collected and the dashboard will be populated in a future release.

When to use

• Diagnosing replication lag between primary and replicas
• Monitoring replication slot growth to prevent WAL retention issues
• Validating HA posture after failover or configuration changes

Key panels (planned)

• Replication lag (seconds and bytes) — delay between primary and each replica
• Slot activity and retention — active vs inactive slots, WAL retained per slot
• Replica state and sync status — streaming, applying, or disconnected

What good looks like

• Replication lag is near zero or within your SLA
• No inactive replication slots growing unbounded (these retain WAL and can fill disk)
• All replicas are in streaming state

What to investigate

Signal	Next step
Growing replication lag	Check replica load, network bandwidth, and max_wal_senders
Inactive slot retaining WAL	Drop unused slots or investigate why the consumer disconnected
Replica not streaming	Check pg_stat_replication on primary and replica logs

Related Checkup checks

• A004 — cluster information