What problem does it solve?
Adaptive Metrics helps reduce Grafana Cloud billing and Prometheus storage by lowering Active Series counts through automated aggregation and label-dropping rules. It addresses high-cardinality metrics that inflate series counts but are not necessary for queries, dashboards, or alerts.
Core Features & Use Cases
- Automatic recommendations: Analyze query history to identify labels that are never queried and propose aggregation rules that drop those labels.
- Rule management: Create, review, and apply aggregation rules (exact or regex matches) via the UI or API to pre-aggregate high-cardinality metrics.
- Cost and safety analysis: Estimate projected series reduction and billing impact, and provide guidance for confirming dashboard, alert, and recording-rule usage before dropping metrics.
- Use case: An SRE team reduces Active Series for a noisy service by applying aggregated rules that drop pod-specific labels, cutting billable series without affecting PromQL dashboards.
Quick Start
Open Grafana Cloud Adaptive Metrics, run a recommendations scan for your desired lookback period, review the top suggestions, and apply the aggregation rules that safely drop unused labels.
