Istio, We Have a Problem! Understanding and Fixing Service Mesh Bugs
summary
Shows diagnosing a broken microservice using Istio: Envoy sidecar captures inter-service calls, Service Graph reveals call rates, and Prometheus/Grafana surface per-service error spikes (one at 50%). Zipkin-style traces identify a failing 24-span trace, then Istio routing (traffic mirroring and canary) and a gradual blue‑green cutover are used to verify and roll out the fix without changing application code.
AI-written summary · about
chapters (12)
- 0:00Istio intro
- 3:10App problem
- 6:59Istio deployment
- 11:24Grafana/Prometheus errors
- 15:02Jaeger traces
- 19:16Bug fix
- 20:40Canary testing
- 25:15Traffic mirroring
- 33:10Blue-green deployment
- 36:50Features summary
- 37:52Installation & compatibility
- 43:05Q&A
related talks
36mIstio, we have a problem! Understanding and fixing bugs with a service-mesh
19mIstio, we have a problem! Understanding and fixing bugs with a service-mesh
28mService Mesh: Demystifying Istio
26mAn Introduction to Service Mesh in Action
25mIstio for Node Developers
52mEmpower Your Microservices with Istio Service Mesh
33mMircoservices the right way: an introduction to Istio
45mIstio to the Rescue