Kubernetes Reloaded: AKA How to Avoid the Dev Plateau with CRDs
summary
Thesis: CRDs stop the “dev plateau” by adding real features you can deploy today. The talk explains the three-step CRD pattern — define CRD, run a controller, create instances — and shows concrete tools: KEDA maps Kafka topic size to an autoscaler (solving HPA’s missing app‑metric scaling), cert‑manager for automated TLS, Kyverno for policies, Crossplane for cloud resources, OpenTelemetry and Grafana operators for telemetry. Live demo swaps a Deployment for Argo Rollouts so Argo Rollouts performs blue‑green across 10 replicas with 20s preview and 30s rollback, and it works.
AI-written summary · about
chapters (12)
- 0:00Kubernetes CRDs intro
- 2:12CRDs as Kubernetes extensions
- 3:58Kubernetes evolution
- 6:25Kubernetes core limits
- 7:28Custom Resource Definitions
- 9:14Create custom resource
- 10:40CRD tools for real problems
- 16:26Grafana with CRDs
- 17:51Crossplane and CRDs
- 19:57CNCF ecosystem with CRDs
- 20:19Demo: Blue-green rollouts
- 28:00CRD future trends
related talks
29mKubernetes Reloaded AKA How to Avoid the Dev Plateau with CRDs
35mLeveling Up Your CD: Unlocking Progressive Delivery on Kubernetes
19mKeynote: CRDs, No Longer 2nd Class Thing!
34mExtending Kubernetes 101
27mDemystifying Why the World Is Built on Kubernetes: Learning To Leverage Bespoke CRDs and Controllers
38mTo Crd, or Not to Crd, That is the Question
19mKeynote: CRDs, No Longer 2nd Class Thing!
48mKubernetes as backend: CRD & Operators