From "Works on My Machine" to "Breaks in Production, But Now We Can Watch"

From "Works on My Machine" to "Breaks in Production, But Now We Can Watch": Champs Journey with OpenTelemetry & SigNoz

You deployed to production. Something is slow. A customer reports an issue. You open the logs and find 47,000 entries that say "Operation completed successfully." Helpful...

In this talk, we'll share how we're instrumenting a multi-tenant .NET 8 platform with OpenTelemetry, shipping traces, metrics, and logs to SigNoz — and how we went from "I have no idea what just happened" to "I have a pretty good idea what just happened, most of the time, if the sampling gods are on our side.”

We'll cover the patterns that actually helped: propagating traces across our message bus so async workflows stop being unsolved mysteries, enriching spans with tenant context so we can tell whose problem it is, custom MongoDB and OpenSearch instrumentation because the out-of-the-box stuff wasn't cutting it, and bridging frontend Sentry traces with backend OTEL spans — because apparently one observability tool is never enough.

We'll also be honest about what we still don't see, the config that's hardcoded when it shouldn't be, and the instrumentation we commented out and never came back to.

No polished "and then everything was perfect" narrative — just a real team trying to understand what their software actually does in production.

The speaker is our very famous Nicolai Oksen, CTO & Partner in Champ.

Programme:
17:00 Welcome and settle in
17:15 Part one
17:45 Sandwich and friendmaking
18:15 Part two
19:00 Bye

There are plenty of parking spots near the venue. Pick one that says Oxygen.

Please sign up no later than at noon the day before. But why wait?