The Phantom Memory Limit
A multi-container Pod has one container stuck in CrashLoopBackOff. The Deployment sets a generous pod-level memory limit, yet raising it changes nothing. Find which container is dying, work out where its real limit comes from, and bring the whole Pod to a stable Running/Ready state.
A Deployment named webstack is running, but its Pod never becomes fully Ready
and keeps crashlooping.
Get it to a stable Running/Ready state.
kubectl get pods -l app=webstack
Hint 1
The Pod has two containers. Only one of them is the problem.
Hint 2
Remember that requests and limits can be set at the Pod level, not only per container. They are not the same thing.
Hint 3
Check the events now and after every change you make.
kubectl events
If you change something and no new pod appears, an event explains why the ReplicaSet could not create it.