GreptimeDB is an observability database for metrics, logs and traces. It provides SQL and PromQL query interfaces and can run as a standalone process or a distributed cluster. Distributed deployments can store persistent data files in shared object storage, separating that storage from the compute serving reads and writes.
Where it fits
Consider GreptimeDB when you want to keep telemetry on infrastructure you control and query different signals through a common database. A practical trial starts with an existing Prometheus or OpenTelemetry data source, realistic retention and representative queries. Compare the result with your current telemetry backend rather than relying on a vendor benchmark.
The core is Apache-2.0 licensed. Enterprise-only features have a separate licence, so check the edition boundary against the operational and security capabilities your team needs.
What to evaluate
Object storage does not remove the need to manage state. The write-ahead log, metadata service, region placement and local caches still affect durability, availability and recovery. Test ingestion bursts, query latency, failure recovery and storage costs with your own data. The architecture guide and user guide explain the components and supported ingestion paths.
