Case Study: Tesla builds a quadrillion-scale observability platform with ClickHouse

How Tesla built a quadrillion-scale observability platform on ClickHouse

Tesla needed a new observability platform that could handle massive, real-time metric volumes across factories, energy systems, and connected vehicles. The company wanted years of retention, high availability, fast queries, and full compatibility with PromQL and existing dashboards and alerts, so it turned to ClickHouse as the data platform for a replacement built at Tesla scale.

Using ClickHouse, Tesla built Comet, a metrics system with scalable ingestion and a transpiler that converts PromQL into ClickHouse SQL while preserving Prometheus API behavior. The result was a platform ingesting tens of millions of rows per second, storing tens of trillions of samples, and proving its resilience with a stress test of over one quadrillion rows ingested in 11 days with flat CPU and memory usage.

Tesla

Alon Tal

Senior Staff Software Engineer