Tempo

Tempo is a distributed tracing backend that ingests OpenTelemetry traces from instrumented services, stores them in object storage, and serves them back for analysis in Grafana through TraceQL queries and trace-to-logs navigation. It is designed to operate as part of the Grafana stack, so traces, metrics, and logs can be correlated without moving data to a separate vendor, and it scales horizontally across collectors.

Teams use it to follow a single request across many services and find the slow span.

Running it on your own infrastructure keeps trace payloads, which often embed request details, on your storage, avoids ingestion based pricing, and makes the block storage easy to back up.