Platform Engineering: Raise the Level of Abstraction using Kubernetes Operators
summary
Shows a working Platform Orchestrator built on Kubernetes Operators and Custom Resource Definitions using the Java Operator SDK and Fabric8, with four CRDs (microservice, backend, frontend, database). Creating an "orders" or "payments" microservice triggers automated Postgres schema creation, user and secret provisioning, Kubernetes deployments and services, and status reconciliation. Deleting a microservice runs cleanup, killing a deployment is self-healed, and operator restarts resume reconciliation, proving the design actually works in the live demo.
AI-written summary · about
chapters (8)
- 0:00Platform engineering intro
- 3:43Team topologies
- 5:30Abstraction levels
- 9:03Kubernetes orchestrator
- 18:15Microservice abstraction demo
- 37:31Data product abstractions
- 39:58Orchestrator patterns
- 43:55Key takeaways
related talks
01h 33mBuilding for Global Engineering Platforms - Let's make an Operator
31mKubernetes Operators within Reach of Any Developer
52mImplementing Microservices as Kubernetes Operators
43mSOS: Simplifying Operators Substantially - A Hands-on Workshop for Beginners in k8s Native Applications
32mA Kubernetes Operator Story: From Idea to Cloud Integration
24mAutomating Stateful Apps with Kubernetes Operators
53mAutomating Stateful Applications with Kubernetes Operators
14mDemystifying k8s operators