PostgreSQL Replication Metrics Guide - Monitoring & Optimization

Replication Slots Information

• What it measures: Tracks the status, activity, and configuration of PostgreSQL replication slots, including their names, plugins, slot types, active status, and restart LSN positions.
• Why it matters: Replication slots are crucial for ensuring that WAL (Write-Ahead Log) data is properly retained and streamed to replicas. Active slots prevent WAL files from being removed before they're received by replicas, ensuring data consistency across your database cluster. Inactive or misconfigured slots can lead to WAL accumulation, storage issues, or replication failures.
• Ideal value & Best Practice:
  • Ensure all replication slots are active and functioning properly
  • Monitor for slot inactivity or stalled replication
  • Regularly validate that active status is true for all production slots
  • Set up alerts for slot inactivity or replication delays
  • Remove unused replication slots to prevent WAL accumulation
  • Monitor restart_lsn advancement to ensure continuous replication
  • Use pg_replication_slots view for comprehensive slot monitoring

Replication Lag Information

• What it measures: Tracks the time and byte difference between primary and replica databases, measuring how far behind replicas are in applying WAL data.
• Why it matters: Minimal replication lag is essential for maintaining data consistency, supporting read scalability, and ensuring quick failover capabilities. Excessive lag indicates performance issues, network problems, or replica overload, which can compromise data freshness and system reliability.
• Ideal value & Best Practice:
  • Aim for replication lag under 100 milliseconds for most applications
  • Monitor both byte lag (pg_wal_lsn_diff) and time lag (reply_time)
  • Set up alerts for lag exceeding acceptable thresholds
  • Investigate sudden increases in replication lag immediately
  • Optimize network throughput between primary and replica nodes
  • Monitor replica server performance and resource utilization
  • Use pg_stat_replication view for real-time lag monitoring
  • Consider using synchronous replication for critical data where zero data loss is required

---

Try pghealth Free Today 🚀

Start your journey toward a healthier PostgreSQL with pghealth.
You can explore all features immediately with a free trial — no installation required.

👉 Start Free Trial