Thanos
Thanos is an open source, CNCF project that extends Prometheus with long-term storage, a global query view across multiple clusters, and high availability for large-scale metrics monitoring.
Definition
Thanos is an open source, CNCF project that extends Prometheus with long-term storage, a global query view across multiple clusters, and high availability for large-scale metrics monitoring.
Overview
Prometheus was designed as a single-server monitoring system, which is simple and reliable but runs into limits at large scale: local disk storage isn't meant for years of retention, and each Prometheus instance can only see the metrics it collected itself. Thanos was created to remove those limits without replacing Prometheus, instead layering on top of it. Thanos components run alongside existing Prometheus servers to ship metric data to cheap object storage (like S3-compatible storage) for long-term retention, while a query layer lets users run a single query across many Prometheus instances and clusters as if they were one system. This solves both the retention problem and the multi-cluster visibility problem that come up as organizations scale Kubernetes deployments across regions or teams. Thanos originated at the gaming infrastructure company Improbable and was later donated to the Cloud Native Computing Foundation (CNCF), where it's maintained as an open source project alongside related tools like Cortex (monitoring) and VictoriaMetrics, which solve similar long-term-storage and multi-cluster problems with different architectural approaches.
Key Features
- Long-term metric storage by offloading Prometheus data to object storage
- Global query view federating metrics across multiple Prometheus instances and clusters
- Deduplication of metrics from redundant, highly available Prometheus pairs
- Downsampling for efficient long-range historical queries
- Compatible with existing Prometheus deployments without requiring a rewrite
- CNCF-hosted open source project with an active contributor community