Documentation

Distributed tracing

Follow a single request as it fans out across your services.

Tracing lets you follow one request end-to-end across every service it touches, so you can see exactly where latency or errors were introduced.

Reading a trace

Each trace is broken into spans representing work done by a single service. Spans are nested to reflect the call graph, and slow or failing spans are highlighted so you can jump straight to the problem.

From incident to trace

Traces referenced by a correlated incident are linked directly from the incident view, so you can pivot from "what broke" to "exactly where it broke" in one click.