Building Reliability into Kubernetes using Chaos Experiments with Terraform
summary
Claims that chaos experiments can be fully automated to improve Kubernetes reliability by provisioning infra with Terraform automation on AWS, installing clusters with kops, and driving failures via scripts (example: a delete-one-pod test and killing master or frontend pods). Demo provisions a 3‑node cluster, runs a script that kills pods, and shows kubectl/dashboard auto‑recreation of replicas — it works. Cites real incidents (Tinder) and large deployments (Buffer: ~60 nodes, ~1,500 containers) to justify running controlled blast‑radius experiments.
AI-written summary · about
chapters (9)
- 0:00Introduction and welcome
- 1:06Reliability in software systems
- 4:16Kubernetes basics and challenges
- 9:20Broken promises of Kubernetes
- 13:56Chaos engineering principles
- 19:14Chaos experiments for Kubernetes
- 23:29Terraform for Kubernetes infrastructure
- 30:42Demo and Kubernetes resilience
- 31:48Q&A and closing
related talks
23mChaos Engineering for Cloud Native
40mJourney into the World of Chaos Engineering
32mImplementing Chaos Engineering in Kubernetes
28mContaining Chaos with Kubernetes
29mImprove Resilience with automated chaos engineering
29mCase Study: Bringing Chaos Engineering to the Cloud Native Develo...
39mPracticing Cloud-Native Resliency Is a Challenge: Decoding It with Chaos...
19mChaos Engineering at Scale