Unified Chaos Injection And SLO Validation Experiments In Kubernetes
Talkat KubeCon + CloudNativeCon North America 2022 · on Thursday, 2022-10-27· at 17:25· Ambassador Ballroom (Room 360)·
summary
Demonstration shows a unified workflow that launches a Helm-packaged experiment which runs load generation, collects metrics and validates SLOs while Litmus injects chaos. It uses iterate CLI to parameterize tests and fetch Prometheus or local metrics, and the litmus chaos engine to run a pod kill every 5s for 1 hour scenario. A single‑pod deployment breaks SLOs (99th percentile ≈3,000 ms), scaling to 2–3 replicas restores SLO compliance.
AI-written summary · about
chapters (5)
- 0:00Start
- 1:11Experiment Examples
- 4:23Load Test HTTP Experiment
- 11:34Load Test gRPC Experiment
- 12:24Resiliency Experiment
related talks
26mCanonical chaos through Litmus
39mPracticing Cloud-Native Resliency Is a Challenge: Decoding It with Chaos...
26mImproving resilience through SLO validation using chaos engineering
28mChaos Validation Made Easy
27mPutting Chaos Into Continuous Delivery to Increase App...
30mBringing Order to Chaos: Make Your Systems More Resilient with Chaos Engineering
23mChaos Engineering for Cloud Native
32mChaos Engineering with Litmus and Jenkins