Kubernetes Reloaded AKA How to Avoid the Dev Plateau with CRDs
summary
Argues that CRDs break the “dev plateau” by letting teams extend Kubernetes with real, reusable primitives and shows working examples: a Kafka-driven autoscaler built with KEDA, certificate lifecycle handled by cert-manager, and a blue‑green demo using Argo Rollouts on a 10‑pod deployment that switches preview → active instantly and can revert immediately. Describes using ArgoCD, Grafana operator and OpenTelemetry collectors as CRDs, shows kubectl/rollout commands, and reports the demos succeed without custom controller code.
AI-written summary · about
chapters (16)
- 0:00Kubernetes CRDs introduction
- 1:08Kubernetes evolving ecosystem
- 3:17Enterprise Kubernetes challenges
- 6:35Built-in feature limitations
- 8:38Custom Resources solution
- 10:51Scaling with CRDs
- 11:53Certificates and secrets
- 12:35Monitoring and alerting
- 13:19Argo CD deployments
- 13:40KubeVirt policy enforcement
- 14:22OpenTelemetry observability
- 14:44Grafana day-two ops
- 15:25Advanced deployment strategies
- 16:07Extending Kubernetes
- 17:53Demo: Blue-green with Argo Rollouts
- 25:24Conclusion
related talks
29mKubernetes Reloaded: AKA How to Avoid the Dev Plateau with CRDs
35mLeveling Up Your CD: Unlocking Progressive Delivery on Kubernetes
34mExtending Kubernetes 101
46mExtend Your Kubernetes With the Power of Open Source
19mKeynote: CRDs, No Longer 2nd Class Thing!
48mKubernetes as backend: CRD & Operators
26mWriting Kubernetes Controllers for CRDs: Challenges, Approaches and Solutions
40mProgressive Delivery with Kubernetes and Argo Rollouts