Chaos Engineering Primer
summary
Describes actively injecting failures into production to prove resilience, with live demos that trigger EC2 terminations and show automated recovery: chaos monkey kills a chosen percentage of instances, chaos lemur terminates nodes from a 40-VM Pivotal Cloud Foundry platform while the health manager auto-restarts them, and locust.io simulates thousands of users to reveal scaling bottlenecks; claims 92% of catastrophic faults stem from unknown misconfigurations and recommends hypothesis-driven steady‑state tests, circuit breakers and autoscalers.
AI-written summary · about
chapters (28)
- 0:00Intro
- 0:41what is an ARCHITECT
- 1:22Risk management
- 2:50A new way to look at organizations
- 4:51Blueprint for living in a Black Swan world.
- 5:44computing
- 8:46Distributed Systems Complexity
- 11:40Chaos Engineering
- 13:08Some outages in the Region
- 13:55Backups
- 14:33Netflix Simian Army
- 15:04Chaos Monkey
- 18:04Other Monkeys
- 23:43Principles of Chaos
- 25:24Hypothesize
- 26:50Vary Events
- 27:30Experiment
- 29:44Automate
- 31:43Reference Architecture for Cloud Native Platform
- 32:24Pivotal Cloud Foundry
- 33:07Chaos Lemur demo
- 36:34Locust demo
- 41:48Lessons Learned
- 44:17The importance of reliability
- 49:10Clean your process
- 50:50Testing Pyramid
- 52:12Further Reading
- 52:57Questions
related talks
18mChaos Engineering: Break It On Purpose
52mGetting started with Chaos Engineering
01h 02mPractical Chaos Engineering breaking things on purpose to make them more resilient against failure
37mChaos Engineering: The Fine Art of Breaking Stuff in Production on Purpose
46mDesigning a System for Chaos
38mChaos Engineering
36mChaos Engineering
48mChaos Engineering: The Path to Reliability