Zipkin

Over its time in service Zipkin proved to be a very unstable solution, at least in combination with ELK. We spent way too much time regularly bringing it back to life.

Picked it as our main distributed tracing system: it gets along with OpenTelemetry out of the box, unlike Jaeger, which still needs a separate Collector.

Zipkin is an open-source distributed tracing system: it follows a request's journey using unique traceId and spanId values, showing how much time is spent in each service, and builds a dependency diagram between them.