Use Chaos Testing to Build Resilience in the Software
summary
Chaos engineering presents a five‑step, disciplined process: state a measurable steady state hypothesis, pick targeted real‑world variables, run automated experiments in production with a minimal blast radius, and iterate; Netflix builds Chaos Monkey and a Simian Army tools to kill services and test failover, AWS adds chaos into its Well‑Architected Framework in 2020, and Netflix sees far fewer outages overall though it still records incidents (an ~one‑hour outage August 2023), demonstrating both effectiveness and limits.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:06What is chaos testing
- 6:27History of chaos engineering
- 12:50Five principles
- 22:51Testing in production
- 28:51Known unknowns and scenarios
- 32:45Benefits of chaos testing
- 35:57Summary
related talks
15mChaos - Does It Need to Be Included in Test Strategy?
41mHow Netflix Embraces Failure to Improve Resilience and Maximize Availability
54mControlled Chaos: The Inevitable Marriage of DevOps & Security
40mResilience Testing - Failure is an Option
40mNavigating the Unknowable: Resilience through Security Chaos Engineering
32mAre you hesitant to run chaos experiment...
27mExecuting Chaos Engineering in Production at a Critical Financial Institution
01h 01mEmbrace chaos to achieve stability