Lightning Talk: Zombie Kubernetes!: Making Nodes Rise From The Dead
summary
Argues that Kubernetes fleets should be provisioned immutably: build nodes from pre-built images or cloud-config/user-data and never patch in place. Requires declarative node manifests (kubelet, kube-proxy, Docker, networking) and automated health checks so the system kills a node and replaces it when a component fails. Masters are harder but feasible with separate etcd clusters and the same replacement pattern; avoids single-master setups. Qbox.io experience and Supergiant tooling inform the approach.
AI-written summary · about
chapters (6)
- 0:00Start
- 0:28Zombie Kubernetes
- 0:49What Is a Mutable Infrastructure
- 1:02Immutable Infrastructure
- 3:40What Do We Need for a Declarative Node Configuration
- 5:33Conclusion
related talks
06mLightning Talk - When Kubernetes is Not the Solution
06mLightning Talk: Beyond Operators: Reimagine Distributed Applications on Kubernetes
07mLightning Talk - Say what? You're Running the Storage Platform IN Kubernetes?
06mLightning Talk: K0s at the Edge: Minimal Kubernetes, Maximum Impact
05mLightning Talk: 7 Tips for Tricks to Enjoy Your Kubernetes Journey
11mLightning Talk: How Jenga & Tetris Can Help Run More Containers with Lesser Costs & Downtime
07mLightning Talk – Bazel & Kubernetes: Develop like it's production
27mKubernetes The Fun Way