Comprehensive Container-Based Service Monitoring with Kubernetes and Istio
summary
Shows how Istio service mesh + Kubernetes wires telemetry to monitoring without code changes, and how a Circonus mixer adapter (about ~20 lines of handler logic) records request counts, errors and log-linear histogram distributions that can be merged across pods. Builds a RED dashboard with a histogram heatmap and inverse-percentile math to answer business questions — e.g. identifying ~4 million requests and two extra ~2 million-request slowdowns that violated an SLI — and to decide if checkout is under or over-provisioned.
AI-written summary · about
chapters (20)
- 0:00Intro
- 2:32Istio Architecture
- 4:16Istio Sample App
- 9:28Service Level Objectives
- 11:39Emerging Standards
- 12:57Containers?
- 16:21Errors
- 19:30WE CAN DO BETTER
- 21:38Log linear histogram
- 23:04Duration - Histogram
- 24:00Duration - SLI
- 25:25Duration - Modes
- 27:03Duration - Heatmap
- 29:45RED - SLI Alerting
- 31:01Istio Metrics Adapter
- 39:48Buffer metrics, then report
- 40:41Your boss wants to know
- 41:39The Slowdown
- 46:21Latency Bands
- 48:00Conclusions
related talks
24mObservability for Istio Service Mesh
52mEmpower Your Microservices with Istio Service Mesh
33mObserving and Troubleshooting your Microservices with Istio
56mObservability in Action for your Service Mesh
33mMircoservices the right way: an introduction to Istio
28mService Mesh: Demystifying Istio
58mUsing Service Meshes and Kubernetes to Solve Service-to-Service Communications...
28mIstio in Action: Taking Kubernetes Traffic Control to the Next Level