Why SREs can't afford to NOT do Chaos Engineering
summary
Claims Chaos Engineering is a scientific method to build measurable confidence and works at any scale. First, attach strace and inject write syscall failures every other write into a 49KB legacy binary; baseline is ~1700 RPS and the experiment shows retries hide failures but slow throughput. Second, clone deployment and add 200ms latency via Toxiproxy on Kubernetes using PowerfulSeal against Goldfinger; heatmap crosses the 100ms threshold and pods are cleaned up automatically. Tools: PowerfulSeal, Toxiproxy, strace, Pumba, Chaos Toolkit, Byteman.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:52Agenda
- 2:08Chaos Engineering
- 5:44Top 5 Myths
- 14:00Scientific Approach
- 18:13Demo
- 23:33Complicated
- 29:18Buyin
- 32:19Shark example
- 34:50Tools
- 37:13Book
- 38:34Hiring
related talks
35mChaos Engineering WG Deep Dive
46mChaos Experiments - Datacenter Stress Testing
24mFreedom of Kubernetes requires Chaos Engineering to shine in production
29mCase Study: Bringing Chaos Engineering to the Cloud Native Develo...
37mUse Chaos Testing to Build Resilience in the Software
15mChaos - Does It Need to Be Included in Test Strategy?
34mChaos Engineering For Hybrid Targets With LitmusChaos
27mPutting Chaos Into Continuous Delivery to Increase App...