Introducing Kubernetes Resource Orchestrator (KRO)
summary
KRO exposes a declarative resource graph definition (RGD) API that generates native Kubernetes objects (deployment, service, ingress) plus cloud CRDs (KCC/ACK/ASO) so a single small YAML can provision a WordPress site with Cloud SQL, PVC and bucket; controllers reconcile state and uses etcd for state, track status fields, and preserve dependency order during reconcile. Project is experimental (v0.4) aiming for v1, vendor-collaborative (Google/AWS/Microsoft); multi-cloud auth and provider upgrade semantics (delete vs in-place upgrade) remain operational constraints.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 1:03Kubernetes architecture
- 3:09Custom Resource Definitions
- 8:51Crossplane overview
- 10:14Introducing KRO
- 13:24Resource Graph Definitions
- 16:59WordPress with cloud SQL
- 20:11Benefits for platform engineers
- 23:18KRO roadmap
- 24:44Q&A
- 26:06Q&A: Deletion order
- 26:48Q&A: Multi-cloud support
- 28:09Q&A: Standalone clusters
- 28:29Q&A: KRO vs Terraform
- 31:16Q&A: State management
- 33:01Q&A: KRO vs Cubilla
- 33:44Q&A: Finalizers
related talks
34mGitOps and KRO: A New Way to Control Cloud Infra
24mKubernetes-Powered Platform Engineering: From IaC To API-Driven Infrastructure
01h 19mRook: Cloud-Native Storage Orchestr...
36mRook Project Intro
29mKubernetes Custom Resources with Kubeless and Metacontroller
34mExtending Kubernetes 101
34mGitOps for the Free Cloud: Bridging Kubernetes and OpenStack with ORC
32mKubernetes as a platform for platforms