From Zero to Billions of Traces: Distributed Tracing at Winamax
summary
Winamax describes building an entirely Open Source observability pipeline (OpenTelemetry collectors, Quickwit, Jaeger) to ingest and query billions of distributed traces while meeting strict player-privacy constraints that ruled out SaaS. The talk explains scaling OpenTelemetry collectors for massive, bursty Champions League traffic, choosing Quickwit to store and query terabytes on object storage, and the implementation strategies (including Node.js auto-instrumentation) used to achieve full coverage in months.
AI-written summary · about
chapters (15)
- 0:00Winamax context
- 1:06Speakers and goals
- 3:35Tech and org architecture
- 5:20Distributed tracing defined
- 9:11Four tracing pillars
- 9:53Traces with OpenTelemetry
- 13:20Trace collection at scale
- 17:29Trace storage with Quickwit
- 24:34Trace search with Jaeger
- 27:23Gradual deployment lessons
- 32:05Jaeger analysis example
- 34:52Infrastructure feedback
- 38:22Team adoption
- 40:51Future improvements
- 41:12Conclusion
related talks
45mMastering Distributed Tracing: Insights from Building a Platform from Scratch
32mDistributed Tracing in Action: Our Journey with OpenTelemetry
01h 02mOpenTelemetry Tracing — Real Challenges in Complex Systems
38mPractical introduction to OpenTelemetry tracing
01h 28mUnderstanding Distributed Systems with OpenTelemetry
26mDistributed Tracing: All the Warning Signs Were Out There!!!
24mDistributed Tracing : New DevOps Foundation.
35mOpenTelemetry for Busy People