Database metrics
Short guide to DB metrics and what they mean.
Last updated: 2026-03-17
This page is a short summary of database metrics. Full details are in:
Where to check first
- Cluster overview panel for current health.
- Metrics panel for trend over time.
- Runtime logs around the same timestamp as spikes.
Common signals
- QPS — queries per second
- IOPS — read/write ops per second
- Disk (MB) — storage used
- Egress (GB) — outbound traffic
Fast interpretation guide
| Pattern | Likely issue | Next action | | --- | --- | --- | | QPS up + latency up | Query contention or under-provisioned plan | Inspect slow queries and scale plan | | IOPS spike + CPU stable | Storage pressure | Check write-heavy jobs and index usage | | Disk linear growth | Retention/cleanup issue | Add retention policy and cleanup jobs | | Egress spike | Traffic burst or data-heavy responses | Verify endpoint behavior and caching |
Troubleshooting
Metrics show "No usage"
- Wait for 1–2 samples to arrive.
- Confirm traffic/queries are actually hitting the cluster.