Use this dashboard to monitor your CloudNativePG clusters across cluster health, connections, throughput, cache and block I/O, replication, WAL and archiving, checkpoints, and storage.
CloudNativePG exposes native cnpg_* metrics on each instance metrics endpoint, and the OpenTelemetry Collector scrapes them for this dashboard.
Dashboards → + New dashboard → Import JSON
Dashboard Coverage
Use these panels to:
- Confirm cluster health: See whether instances are up, the PostgreSQL version, replica count, and whether a switchover or fencing is in effect.
- Track connections: Monitor backends by state and database, and backends waiting on locks.
- Measure throughput: Track transactions, tuple activity, deadlocks, conflicts, and temp file usage.
- Inspect cache efficiency: Watch the buffer cache hit ratio and block I/O rates and timing.
- Watch replication: Track standby lag, primary write/flush/replay lag, replication status, and slot lag.
- Follow WAL and archiving: Monitor WAL directory size, segment count, archive status, and time since the last archive.
- Review checkpoints and storage: Track checkpoint rate and timing, buffers written, database size, and transaction ID age.
Metrics Included
Cluster Health
- Cluster Up: Whether each instance is up (
cnpg_collector_up). - PostgreSQL Version: Reported PostgreSQL version (
cnpg_collector_postgres_version). - Streaming Replicas: Number of streaming replicas.
- Nodes Used, Replica Mode, Fencing, Switchover Required: Cluster topology and operational state.
Connections & Backends
- Total Backends and Backends by State: Connection counts overall and by state.
- Backends by Database: Connections grouped by database.
- Backends Waiting on Locks: Backends blocked waiting on locks.
- Longest Transaction and Longest Transaction by Database: Maximum transaction duration.
Throughput
- Transactions /s: Commit and rollback rates.
- Tuple Activity /s: Rows fetched, returned, inserted, updated, and deleted.
- Deadlocks & Conflicts /s: Deadlock and recovery conflict rates.
- Temp Files /s and Temp File Bytes /s: Temporary file activity.
Cache & Block I/O
- Buffer Cache Hit Ratio: Share of block reads served from cache.
- Block I/O /s and Block I/O Time /s: Block read and write rates and time spent.
Replication
- Replication Lag (standby): Lag on standby instances (
cnpg_pg_replication_lag). - Write / Flush / Replay Lag (primary view): Per-stage replication lag from the primary.
- Replication Status and Replication Slot Lag: Receiver state and slot WAL lag.
WAL & Archiving
- WAL Directory Size and WAL Segment Count: WAL volume on disk.
- WAL Archive Status and WAL Archiver Ops /s: Archive state and archive/fail rates.
- Time Since Last Archive / Failure: Age of the last successful and failed archive.
Checkpoints
- Checkpoints /s: Timed and requested checkpoint rates.
- Checkpoint Write / Sync Time /s: Time spent writing and syncing during checkpoints.
- Buffers Written /s: Buffers written by the checkpointer and background writer.
Storage & Collector
- Database Size and Total DB Size: Per-database and total size.
- Transaction ID Age and MultiXact ID Age: Wraparound risk indicators.
- Metrics Collection Duration, Metrics Collections /s, Collection Errors: Health of the CloudNativePG metrics collector itself.
Dashboard Variables
Use these filter variables:
- pod: Filter by CloudNativePG instance pod.
- datname: Filter per-database panels by a specific database.
