Achieving End-to-end Visibility into Complex Event-sourcing Transactions using Distributed Tracing
summary
Claims end-to-end visibility for message-driven microservices using OpenTelemetry and shows how to prove it. Builds a Node.js producer and consumer with kafka js, first manually creating spans and injecting trace context into Kafka headers, then replacing that with the kafka-js auto-instrumentation. Traces ship to Jaeger and show the producer span as parent and consumer spans as children, so message flow and timing are visible and multiple consumers correlate correctly.
AI-written summary · about
chapters (9)
- 0:00Start
- 0:52Agenda
- 1:41About me: Michael Haberman
- 2:20End to end visibility
- 4:41We want to see!
- 5:30Traces!
- 9:29How traces works?
- 12:08DEMO
- 25:17What we learned?
related talks
41mTracing event-based systems with OpenTelemetry
26mThe Art of Event Driven Observability with OpenTelemetry
43mTracing Kafka with OpenTelemetry
27mEnd-to-end visibility in the cloud w/OpenTelemetry
38mPractical introduction to OpenTelemetry tracing
51mPractical introduction to OpenTelemetry tracing
49mDistributed Tracing: How the Pros Debug Concurrent and Distributed Systems
47mI'm Tracing You!