ทำงานกับ kubectl: เครื่องมือควบคุมคลัสเตอร์ของคุณ

kubectl is the official CLI that talks to the Kubernetes API server. Nearly every cluster operation, creating, reading, updating, deleting resources, goes through it.

kubectl reads a kubeconfig file (default ~/.kube/config) that holds cluster endpoints, users, and credentials grouped into contexts.

kubectl get lists resources. Add -o wide for more columns or -A for all namespaces.

kubectl describe shows detailed, human-readable status plus an events section that is invaluable for debugging.

Apply a manifest file with kubectl apply -f. This is declarative: you describe the desired state and Kubernetes reconciles to it.

For quick experiments you can create resources imperatively.

Stream container output with kubectl logs. Use -f to follow and --previous to see a crashed container.

Open a shell inside a running container to debug interactively.

Edit live resources or remove them. Prefer editing manifests in git over edit for production.

Most commands accept -n namespace. Set a default namespace to avoid repeating it.

Use -o yaml or -o json to see the full object, and -o jsonpath to extract a single field.

Test what you have learned.

You practiced the core kubectl verbs: get, describe, apply, logs, exec, delete, plus contexts, namespaces, and output formats to control and debug your cluster.