Cracking Microservices Performance Issues with Distributed Tracing
summary
Distributed tracing demo uses Jaeger with OpenTelemetry and a Hot Rod sample app to find concrete bottlenecks: a customer-service span shows a 1.18s lock wait out of 1.45s total, driver calls issue a staircase of sequential Redis requests (169ms total, longest 32ms) that could be parallelized, and Jaeger reconstructs live service graphs. Survey numbers appear: 47% adoption and Jaeger at 32% share, sampling defaults near 0.1%.
AI-written summary · about
chapters (11)
- 0:00Introduction and agenda
- 1:13Why distributed tracing matters
- 6:34Basic terms and concepts
- 12:42Jaeger overview and history
- 13:46Live demo of Jaeger
- 21:51Demo summary and best practices
- 24:00Instrumentation with OpenTelemetry
- 26:31Common performance patterns
- 28:58Future of distributed tracing
- 30:46Summary and key takeaways
- 32:13Q&A session
related talks
37mOpenTelemetry & Distributed Tracing: A Jaeger v2 Deep Dive
34mWould You Like Some Tracing With Your Monitoring?
32mInvestigating Performance Issues In Microservices Arch.
35mJaeger Intro
37mDeep Dive: Jaeger
51mDistributed Tracing for microservices with Jaeger
31mUnderstanding Microservices with Distributed Tracing
28mIntro: Jaeger