thanos-io/thanos

Thanos is a set of components that adds highly available metric storage and querying to existing Prometheus deployments.

Thanos solves the problem of scaling Prometheus beyond a single instance while retaining historical data indefinitely. It works by leveraging Prometheus's native storage format to store metric blocks in object storage, then providing a unified query interface across multiple Prometheus servers. The system deduplicates metrics from Prometheus HA pairs on the fly and supports downsampling of historical data to accelerate queries. It can be deployed as a sidecar alongside Prometheus or as a separate receive component to handle remote write traffic, making it adaptable to different infrastructure patterns.

Teams should adopt Thanos when they need to query metrics across multiple Prometheus instances, require long-term metric retention beyond what a single Prometheus server can handle, or want to eliminate single points of failure in their monitoring stack. The tool suits organizations already running Prometheus who want to extend it without replacing their existing setup. It requires only object storage as an optional dependency, supporting any S3-compatible system or Google Cloud Storage, which keeps operational overhead minimal.

The project maintains a substantial base of real-world adopters, as evidenced by the fact that nearly all open issues come from outside users rather than the core team. Responses to issues and pull requests typically arrive within one to two weeks. Development activity concentrates on bug fixes, feature requests and improvements, with particular focus on the query component.