Topolearnkubectl is the tool you will type a thousand times a day; its patterns are worth learning once.
kubectl reads connection details from a kubeconfig file (~/.kube/config, or the KUBECONFIG variable). It holds clusters (API URL, CA), users (credentials) and contexts that pair a cluster, a user and a default namespace.
kubectl config get-contextslists them;kubectl config use-context prodswitches.kubectl config set-context --current --namespace=shopchanges the default namespace.
Commands follow kubectl <verb> <resource> [name] [flags]:
get pods,get deploy,svc,get pods -o wide,-o yamlto see the full object,-wto watch.describe pod web-123: human-readable detail plus recent events, the first place to look when something is wrong.apply -f file.yaml: create or update from a file or directory.delete -f file.yamlordelete pod web-123.-n nstargets a namespace,-Aall namespaces,-l app=webfilters by label.
kubectl explain deployment.spec.strategy prints the schema documentation for any field, straight from the cluster. kubectl api-resources lists every kind and its short name (po, deploy, svc, cm).
Always know which context you're in before running a destructive command: many people show it in their shell prompt.
Example
kubectl config get-contexts
# CURRENT NAME CLUSTER NAMESPACE
# * kind-dev kind-dev default
# prod prod-eks shop
kubectl get deploy,svc -n kube-system
kubectl get pods -l app=web -o wide
kubectl get pod web-6d4cf56db6-7qv9z -o yaml | head -20
kubectl describe pod web-6d4cf56db6-7qv9z # read the Events at the bottom
kubectl explain pod.spec.containers.ports
kubectl api-resources | grep -i deploy
# deployments deploy apps/v1 true DeploymentRemember
- A context = cluster + user + default namespace
- describe shows events; -o yaml shows the full object
- kubectl explain documents any field from the live API
Watch out for
- Running kubectl delete or apply against prod because the current context was not the one you thought.