TRACING AKKA STREAMS
summary
Implements Dapper-style distributed tracing using htrace and Zipkin and ships a small library that plugs into Akka Streams and HTTP with a config change only. Demo shows end-to-end traces across front-end and back-end, nine spans observed and parallel work across nine threads after adding async and propagating trace headers. Uses sampling ideas (Google used 1/1024) to avoid overload. Library works in the demo, requires adding a tracing dispatcher or headers to propagate traces.
AI-written summary · about
chapters (8)
- 0:00Tracing matters for Akka Streams
- 1:52Distributed tracing fundamentals
- 3:20Google’s Dapper paper overview
- 5:51OpenTracing and Zipkin
- 6:55Tracing Akka Streams with minimal changes
- 7:29Live demo of tracing
- 11:55Tracing HTTP with Akka
- 14:20Key takeaways and resources
related talks
57mDistributed Tracing: How the Pros Debug Concurrent and Distributed Systems
49mDistributed Tracing: How the Pros Debug Concurrent and Distributed Systems
42mAn introduction to distributed tracing and Zipkin
24mDistributed Tracing : New DevOps Foundation.
54mA TRACE TO REMEMBER: COMPOSITIONAL TRACING
45mLatency tracing in distributed Java applications
50mDistributed Tracing in the Wild
20mThe (Full)stack Trace: Understand Your App with Distributed Tracing