CostLensBigQuery Optimization (GCP)
BigQuery Optimization (GCP)
CostLens surfaces BigQuery capacity-commitment recommendations from Google's Recommender to right-size slot commitments to actual usage.
Updated July 20261 min read
CostLens surfaces BigQuery capacity-commitment recommendations from Google's Recommender. Over-committed BigQuery slots bill continuously regardless of use; the Recommender validates the right commitment level from your query patterns.
How it works
Recommender is queried
CostLens reads the BigQuery capacity-commitments cost recommender across the project's regions.
Recommendations are ingested
Each cost recommendation becomes a finding with the projected monthly saving (from the Recommender's cost projection).
What to do
- Review the recommended slot commitment against your query workload and concurrency needs.
- Adjust your capacity commitments / reservations (BigQuery → Capacity management), or move bursty workloads to on-demand.
Severity levels
| Severity | Estimated monthly savings |
|---|---|
| critical | >$500 |
| high | $100–$500 |
| medium | $20–$100 |
| low | <$20 |
Required permissions
- Detect:
roles/recommender.viewer(+ BigQuery read) — see GCP Accounts.
Limitations
- Applies to projects using BigQuery capacity/slot commitments; on-demand-only projects won't have these recommendations.
- Recommendations require query history to generate.