Configure recording rules
Recording rules allows you to periodically pre-compute frequently used or computationally expensive queries, saving the results as a new time series metric.
For instance, you can create a recording rule generating a new metric, error_9001_count
, which counts occurrences of a specific log error within one minute. Then, query the error_9001_count
metric in dashboards and alert rules.
Recording rules can be helpful in various scenarios, such as:
- Faster queries are needed: Performing heavy aggregations or querying large data sets is quicker with precomputed results than real-time queries.
- Reducing system load: Precomputing specific queries in advance can reduce system overload caused by multiple simultaneous queries.
- Simplifying complex aggregations: Create a new metric from complex aggregations to facilitate alert and dashboard setup.
- Reusing queries across alerts: Improve efficiency by reusing the same query across similar alert rules and dashboards.
The evaluation group of the recording rule determines how often the metric is pre-computed.
Similar to alert rules, Grafana supports two types of recording rules:
- Grafana-managed recording rules, which can query any Grafana data source supported by alerting.
- Data source-managed recording rules, which can query Prometheus-based data sources like Mimir or Loki.