Getting started with Chaos Engineering
summary
Chaos Engineering runs controlled failures against production-like systems to prove resilience: here an auto‑scaling group of 20 instances has 10 instances killed over 10 minutes via chaos monkey CLI to verify replacement and service continuity; monitoring collects ~50,000 metric streams (about 4.8 million metrics in seven minutes) and uses tracing with Jaeger. Tools discussed include Chaos Toolkit, Gremlin and in‑house VM‑canary; common tests are instance shutdowns, network latency/security group cuts, and CPU/IO saturation.
AI-written summary · about
chapters (4)
related talks
57mChaos Engineering Primer
46mDesigning a System for Chaos
18mChaos Engineering: Break It On Purpose
36mChaos Engineering
26mBreaking Systems for Resilience
23mHow to tame chaos effectively?
37mChaos Engineering: The Fine Art of Breaking Stuff in Production on Purpose
21mThe Psychology of Chaos Engineering
