Observing chaos: how distributed tracing brings observability to a service mess
summary
Demonstrates reclaiming visibility lost to a service mesh by combining Istio for policy and fault injection with Jaeger tracing and Quarkus auto-instrumentation. Deploys three microservices with three containers per pod (app, sidecar proxy, Jaeger agent), injects a 10s delay and enables mutual TLS, then inspects a live trace that shows 27 spans and a 14 ms gap before the app span. Shows concrete YAML examples for delay and mTLS and that traces let operators pinpoint which hop and span cause failures.
AI-written summary · about
chapters (7)
- 0:00Introduction to concepts
- 1:24Service mesh theory
- 8:37Challenges and benefits
- 16:50Observability and tracing
- 27:45Tooling introduction
- 35:54Live demo and application
- 43:47Conclusion and Q&A
related talks
05mIgnite: Istio+: Enhancing Observability and Tracing using Kiali and Jaeger
20mKeynote: Service Meshes and Observability
42mDistributed Monitoring: How to understand the Chaos
45mTracing Cloud-Native Applications: From stacktrace to Opentracing using Istio
35mNavigating the Distributed Systems Execution Maze With OpenTracing
01h 14mIntro + Deep Dive: Jaeger
39mThe Hidden Gems of Distributed Tracing
31mDistributed Tracing Integration with OpenTelemetry and Knative