CI/CD with Kubernetes - the Good, the Bad and the Ugly
summary
Claims Kubernetes makes CI/CD reproducible and dynamic and demonstrates it by actually creating a namespace, deploying nginx manifests, curling the endpoint and garbage‑collecting the namespace so the pipeline succeeds and cleans up. It uses dynamic namespace creation, job executors as pods and the cluster autoscaler to handle spikes. Pain points are concrete: plugin fragmentation, fragile third‑party integrations, costly unchecked resource consumption and a multi-cluster permissions nightmare that complicates promotions and credentials.
AI-written summary · about
chapters (4)
related talks
36mThe Kubernetes hitchhiker's guide to CI/CD
39mKubernetes-centric CI/CD
23mKeynote: KubeCon Opening Keynote
47mContinuous Delivery with Containers: The Good, Bad, & Ugly
51mKubernetes for the Impatient
40mThe Good, the Bad, and the Ugly of Managing Production Systems with Kubernetes
28mAutomate Everything
56mContinuous Deployment Pipeline in Kubernetes