Architecting Kubernetes-Based Internal Developer Platforms: Essential Patterns...
summary
Explains a concrete Kubernetes-based IDP architecture: an automated Crossplane-driven provisioner plus API/CLI/GUI interfaces, tenancy choices of namespace vs logical-cluster vs full-cluster tenancy, and SSO via OIDC with projects like Dex; recommends dedicated teams to run DB-as-a-service operators and highlights operational costs and compatibility pain of day‑2 ops. Shares hard numbers and experience: platform runs 2,000 production Kubernetes clusters, built a custom provisioner to scale, and emphasizes choosing tools based on platform team skills and automated workflows.
AI-written summary · about
chapters (9)
- 0:00Kubernetes-based IDPs intro
- 2:36What is platform engineering
- 7:51Kubernetes as a service
- 14:48Provisioner and interface
- 19:09X as a service with operators
- 25:19Isolation and multi-tenancy
- 31:26Single sign-on clusters
- 34:23Key takeaways
- 36:12Q&A
related talks
11mBuilding Scalable Developer Platforms
32mLaying the Foundation: Real World Kubernetes Deployment Patterns
31mKubernetes is Your Platform: Design Patterns For Extens...
27mWhy Developer Platforms and a curated Dev Experience are a must have when scaling
19mProduction-Ready Platform Engineering: Build To FAIL
45mDevOps Is All About Building Internal Developer Platform (IDP)
43mPlatform Engineering Redefined: Evolving Kubernetes Beyond Container Orchestration
40mThe Elephant in the Kubernetes Room: Team Interactions