Challenge: diagnose and fix a broken Pod

In the previous lesson you created a Pod that worked and inspected it with kubectl describe and kubectl logs. Now you're going to use those same tools for what they're really for: diagnosing Pods that don't work.

The challenge hands you a cluster with two broken Pods. One is in ImagePullBackOff, the state that appears when the kubelet can't pull the container's image. The other is in CrashLoopBackOff, the loop in which Kubernetes restarts, over and over, a container whose process exits as soon as it starts. They are, by far, the two error states you'll see most often in your life with Kubernetes.

There are no guided steps: you have the problem, the collapsed hints in case you get stuck, and the automatic verification. Before opening a hint, exhaust what describe and logs tell you.

Passing criterion: the Pods api and worker are in the Running state, ready and not accumulating restarts. Completing the challenge passes this lesson.