Kubernetes pod stuck in CrashLoopBackOff, no logs

kubectl logs for my pod returns nothing and the pod restarts every few seconds. How can I see why it crashes?

Look at the previous container and the events:

kubectl logs mypod --previous
kubectl describe pod mypod   # check Last State, Exit Code, Events

Exit code 137 is an OOM kill; 1 is usually the app itself; a missing command shows up in Events.

--previous showed a missing env var. Thanks!