kubectl Debugging Cheatsheet

kubectl get pods -A --field-selector=status.phase!=Running

→ Instantly shows every pod that is not running.

kubectl get pod <name> -o wide

→ Shows node, IP and restart count.

kubectl describe pod <name>

→ The most important command - solves 80 % of cases.

kubectl logs <pod> --previous

→ Returns the logs of the previous crash.

kubectl logs <pod> -f --tail=100

→ Streams the latest logs in real time.

kubectl get events --sort-by=.lastTimestamp -A | tail -30

→ Helps when describe falls short.

kubectl debug -it <pod> --image=busybox --target=<container>

→ Ideal for distroless images or pods without a shell.

kubectl describe node <name>

→ For pending pods, evictions or resource issues.

kubectl top pods -A --sort-by=memory

→ Finds memory hogs and suspicious pods.

kubectl get all -n <ns>

→ Shows every key resource in a namespace.

kubectl exec -it <pod> -- /bin/sh

→ When the container has a shell.

kubectl auth can-i --list -n <ns>

→ Helps with permissions and role issues.