Observability with Prometheus and Grafana

Long-Term and Clustered Storage

Prometheus is not designed for long-term storage. It is optimized for real-time monitoring and alerting, not for storing historical data.

ℹ️ The default retention period for Prometheus is 15 days, which means that it will only store data for the last 15 days. After that, the data is deleted. This can be changed by modifying the --storage.tsdb.retention.time flag in the configuration file.

The size of the Prometheus database is limited by the amount of available disk space. If you need to store data for longer periods, you can use remote storage solutions. Elasticsearch, InfluxDB, Amazon S3, and other integrations are available to extend the storage capabilities of Prometheus.

Clustered storage (data split and replicated across many nodes working together) is not supported by Prometheus. If you need to store data across multiple nodes, you will need to integrate with a tool like Uber's M3DB