10 Kubernetes is hard!
summary
Company migrates CI/CD into Kubernetes, running hundreds of environments per week via a single Jenkins job and shared Helm charts that are also used in production. Tooling includes Prometheus + Grafana + EFK for visibility, fluent-bit for logs, kubectl and k9s for debugging, and Helm for deployment. Early failures (node crashes from missing resource limits, Java and RabbitMQ quirks, Mongo issues) lead to fixes like probes, limits, and helm-driven configs; a full multi-region deploy completes in ~20 minutes.
AI-written summary · about
chapters (12)
- 0:00Introduction and journey
- 3:07Starting small with Kubernetes
- 4:50Preparing applications for Kubernetes
- 7:20Security in containerized environments
- 8:27Kubernetes as a solution
- 9:54Resource limits and management
- 11:42Application health and probes
- 12:46Logging and lifecycle management
- 13:49Helm for deployments
- 16:18Visibility and debugging tools
- 19:09Automated deployments with Jenkins
- 20:54Key takeaways and community
related talks
35mOff the Beaten Path: An Explorer’s Guide to Kubernetes
53mDeploying to Kubernetes Thousands of Times Per/Day
38mScaling Kubernetes Deployments Across +15'000 Namespaces
35mKubernetes journey
50mContinuous Deployment With Jenkins X And Kubernetes
35mTo Infinite Scale and Beyond: Operating Kubernetes Past the Steady State
20mKeynote: The Challenges of Migrating 150+ Microservices to Kubernetes
35mJenkins Scaling and Organization for an Efficient CI