Cluster API Deep Dive With a Tencent Case Study
summary
Explains Cluster API internals and shows a working Tencent provider that uses CRDs and controller deployments plus cloud secret id/secret key to create, upgrade and delete machines. Demonstrates a rolling upgrade flow from v1.8.3→v1.9.3 where MachineDeployment creates new MachineSets and replaces nodes under the adapter, and describes bootstrapping options (in-cluster vs manager cluster). Claims production use managing 2,000+ TKE clusters and reuses community adapters to avoid duplicated cloud-specific code.
AI-written summary · about
chapters (13)
- 0:00Start
- 1:58Where we are
- 4:34What's Cluster API?
- 6:14What is configured?
- 8:44YAML to Objects
- 9:27Objects to Machines
- 11:15Case Study: Upgrade
- 14:40Bootstrapping-10 km view
- 17:37Deploy-In Cluster
- 18:40Deploy-Out of Cluster
- 19:31Provider Support
- 20:09DEMO on TKE
- 34:29Cluster-api for TKE internal
