The Psychology of Chaos Engineering
summary
Argues that chaos engineering improves system and team resilience but also creates human costs if mishandled, illustrated with case studies and concrete practices: failure Fridays as planned daytime experiments, rolling out to a 1% canary and aborting on a key metric such as stream starts per second. Describes tools and artifacts used (Netflix Simian Army, Gremlin, Chaos Toolkit, chaos slinger, internal Cass cat), certification cadence (every ~180 days), blast‑radius limits, clear abort criteria and transparent notifications to protect users and responders.
AI-written summary · about
chapters (22)
- 0:00Introduction
- 1:49Chaos Engineering
- 3:49Is all engineering chaotic
- 4:43Its not about breaking things
- 5:25Experimentation
- 7:02Production
- 9:23Measurement
- 10:33Blast Radius
- 12:42Tools
- 16:54The People
- 17:28Netflix
- 17:55Banking
- 19:21How do you feel
- 20:21Management can get nervous
- 20:31Unfortunate names for things
- 20:59Making transformative change
- 21:39Mental shift
- 22:34Internships
- 23:56Safety Matters
- 24:16Know Your Conditions
- 25:27This Isnt About Stress
- 26:53Resources
related talks
21mThe Psychology of Chaos Engineering
26mPsychology of Chaos Engineering
18mChaos Engineering: Break It On Purpose
27mChaos Engineering: The Path to Reliability
52mChaos engineering: Breaking things on purpose in production
36mChaos Engineering
36mChaos Engineering in 2021
48mChaos Engineering: The Path to Reliability