Troubleshooting
Guides for diagnosing and resolving common PostgresAI monitoring issues.
Quick diagnostics​
Check component status​
# Docker Compose
docker compose ps
# Expected output
# NAME STATUS
# pgwatch Up
# victoriametrics Up
# grafana Up
Verify metrics flow​
# 1. Check pgwatch is collecting
curl http://localhost:8080/metrics | grep pg_stat
# 2. Check VictoriaMetrics is receiving
curl 'http://localhost:8428/api/v1/query?query=up'
# 3. Check Grafana data source
curl http://monitor:YOUR_PASSWORD@localhost:3000/api/datasources/proxy/1/api/v1/query?query=up
Common issues​
| Symptom | Likely cause | Guide |
|---|---|---|
| "No data" in all panels | Collection not running | No data troubleshooting |
| "Access denied" errors | Missing permissions | Permission errors |
| Slow dashboards | Query performance | Performance tuning |
| Missing pg_stat_statements | Extension not loaded | No data troubleshooting |
Diagnostic commands​
pgwatch logs​
docker compose logs pgwatch --tail 100
VictoriaMetrics logs​
docker compose logs victoriametrics --tail 100
Grafana logs​
docker compose logs grafana --tail 100
PostgreSQL connectivity​
docker compose exec pgwatch psql -h target-host -U monitoring_user -c "select 1"
Health check endpoints​
| Component | Endpoint | Expected |
|---|---|---|
| pgwatch | http://localhost:8080/health | {"status": "ok"} |
| VictoriaMetrics | http://localhost:8428/health | OK |
| Grafana | http://localhost:3000/api/health | {"database": "ok"} |
Getting help​
- Check logs for error messages
- Review the specific troubleshooting guide
- Search GitHub Issues
- Open a new issue with diagnostic output
Sections​
- No data troubleshooting — Empty dashboards
- Permission errors — Access denied issues
- Performance tuning — Slow queries and dashboards