Speaker Panel - Day 1
summary
Panel argues chaos engineering works when teams opt-in not opt-out and carefully limit blast radius — one early experiment took down QA for a week. Adoption succeeds with leader buy‑in plus engineer‑to‑engineer trust; automated runners and phased rollouts matter. Common failure modes are vague problem definitions (no shared answer to “what is done”) and missing monitoring, so advice emphasizes start with observability, psychological safety, listening first, and curated reading like Sydney Decker’s field guide and the Resilience Papers Club.
AI-written summary · about
chapters (13)
- 0:00Introduction and panel setup
- 1:04Why holistic resilience matters
- 7:37Cross-industry resilience lessons
- 13:27Chaos engineering in safety-critical systems
- 19:53Overcoming expert resistance
- 23:05Adopting chaos engineering at Netflix
- 28:08Cultural change and psychological safety
- 30:16Problem definition pitfalls
- 33:08Teaching resilience to new engineers
- 41:15Designing for human limitations
- 43:49Sustained adaptive capacity
- 46:27One piece of advice for resilience engineering
- 48:13Closing remarks and resources
