Pod Pending vs CrashLoopBackOff
Pending means the pod never got a node (or never started). CrashLoopBackOff means it started and keeps dying.
Pending
Kubernetes: Pod stuck in Pending — FailedScheduling: 0/3 nodes are available
The scheduler can’t find a node that satisfies the pod’s resource requests, selectors, taints or volume constraints.
Causes & fixes →
CrashLoopBackOff
Kubernetes: CrashLoopBackOff
The container starts, crashes (or exits), and Kubernetes keeps restarting it with increasing back-off delays.
Causes & fixes →| Pending | CrashLoopBackOff | |
|---|---|---|
| Container ran? | No | Yes, repeatedly |
| Look at | kubectl describe pod → Events (FailedScheduling) | kubectl logs --previous + Last State |
| Typical cause | Resources, selectors, taints, PVCs | App crash, bad config, OOM, failing probes |
Rule of thumb
Pending is a scheduling problem; CrashLoopBackOff is an application/runtime problem.