Chaos & Intuition Engineering at Netflix
summary
Netflix runs chaos experiments against its control plane (deployed in three AWS regions) using tools like Chaos Monkey and Chaos Kong to kill servers and simulate regional failures, and reports ~400 microservices and ~4 billion streaming hours/month. Experiments spin up control and experimental clusters, assert a hypothesis about steady‑state behavior, and automate continuous runs; a visual tool called visceral gives an at‑a‑glance health intuition and has already prevented at least one major operational mistake.
AI-written summary · about
chapters (16)
- 0:00Intro
- 0:45What is Netflix
- 3:15Microservice Architecture
- 7:18The Beer Game
- 12:45Bullwhip Pattern
- 14:48Netflix
- 17:48Chaos Monkey
- 21:44Chaos Engineering
- 22:30Testing Resilience
- 24:03Community Day
- 25:07Advanced Principles
- 29:05Intuition Engineering
- 32:21Visceral
- 39:28Microservices
- 40:58Have you ever gotten in trouble
- 43:08How do you preserve the user experience
