Troubleshooting: Start Here¶

Most incidents are solved faster by following the same steps every time than by guessing.

A repeatable method¶

1. Define the symptom. What exactly is failing, for whom, and since when?
2. Check what changed. Deployments, config, node upgrades and certificate expiry cause most outages.
3. Walk the layers from the outside in: DNS, then Service, then endpoints, then Pod, then node.
4. Gather evidence before you change anything: events, logs and describe output.
5. Change one thing at a time, and note what you did.

The five commands I reach for first¶

kubectl get pods -A -o wide
kubectl describe pod <pod> -n <ns>
kubectl logs <pod> -n <ns> --previous
kubectl get events -n <ns> --sort-by=.lastTimestamp
kubectl debug -it <pod> -n <ns> --image=nicolaka/netshoot --target=<container>

describe and events explain most scheduling, image and probe failures. --previous shows the logs of the container that just crashed. kubectl debug attaches a temporary toolbox container to a live Pod.

Which page do I need?¶