CKA v1.35 · Troubleshooting (30%) → "Troubleshoot clusters and nodes" / Workloads — "Configure Pod admission and scheduling" Target time: 8 minutes
Namespace stuck has Deployment api (2 replicas). Its pods have been
Pending for ages and someone says "the cluster is full."
It isn't. Find out what the scheduler actually objects to — there is more than one thing — and fix the Deployment IN PLACE so both replicas run. Do not delete and recreate it, do not add nodes, do not label nodes.