Troubleshooting Pods and Containers

Jun 6 2026

On this page

Debugging playbook

Pod issue: describe → logs → exec

CrashLoopBackOff: logs → probes → command / args

Service issue: selector → endpoints → Pod labels

Pending: describe → Events → scheduling / resources / PVC

Drill-down order

kubectl get po: READY, STATUS, RESTARTS
kubectl describe po: Events section
kubectl get events --sort-by=.metadata.creationTimestamp
kubectl logs / logs -p (previous instance)
kubectl exec -it / -c <container>
kubectl debug (ephemeral container for distroless)
kubectl port-forward po/<n> 8080:80
kubectl top po / kubectl top nodes (metrics-server)

Symptom → cause → debug command

Symptom	Likely Cause	Debug Command
CrashLoopBackOff	App crashes on start	`kubectl logs <pod> --previous`
ImagePullBackOff	Wrong image name/tag or missing registry secret	`kubectl describe pod <name>`
Pending	No node fits (resources/taints/affinity)	`kubectl describe pod <name>` → Events
OOMKilled	Container exceeded memory limit	`kubectl describe pod` → 'OOMKilled'
Init:0/1	Init container not completing	`kubectl logs <pod> -c <init-container>`
0/1 Running	readinessProbe failing	`kubectl describe pod` → probe events

Container logs

Task	Command
Logs from pod (single container)	`kubectl logs <pod>`
Logs from specific container	`kubectl logs <pod> -c <container>`
Logs from crashed/previous container	`kubectl logs <pod> --previous`
Stream logs live	`kubectl logs <pod> -f`
Last 50 lines	`kubectl logs <pod> --tail=50`
Logs since 1 hour ago	`kubectl logs <pod> --since=1h`
Logs from all pods with label	`kubectl logs -l app=webapp --all-containers=true`
Logs from init container	`kubectl logs <pod> -c <init-container-name>`

Monitoring & CLI tools

Task	Command
CPU/memory for pods	`kubectl top pod`
CPU/memory for pods in namespace	`kubectl top pod -n <ns>`
CPU/memory for nodes	`kubectl top node`
Sort pods by CPU	`kubectl top pod --sort-by=cpu`
Sort pods by memory	`kubectl top pod --sort-by=memory`
Recent events (sorted)	`kubectl get events --sort-by='.lastTimestamp'`
Events for specific namespace	`kubectl get events -n <ns>`
Watch events live	`kubectl get events -w`
Pod with node + IP	`kubectl get pod -o wide`
Full pod YAML	`kubectl get pod <name> -o yaml`
Watch pod status	`kubectl get pod -w`
JSONPath: get pod IP	`kubectl get pod <name> -o jsonpath='{.status.podIP}'`
JSONPath: all container images	`kubectl get pod <name> -o jsonpath='{.spec.containers[*].image}'`
Custom columns	`kubectl get pods -o custom-columns=NAME:.metadata.name,NODE:.spec.nodeName`

Debugging in Kubernetes

Task	Command
Shell into running container	`kubectl exec -it <pod> -- /bin/sh`
Shell into specific container	`kubectl exec -it <pod> -c <container> -- bash`
Run one-off command	`kubectl exec <pod> -- cat /etc/config/app.conf`
Debug with ephemeral container	`kubectl debug -it <pod> --image=busybox --target=<container>`
Copy of pod + debug container	`kubectl debug <pod> -it --image=busybox --copy-to=debug-pod`
Copy file from pod	`kubectl cp <pod>:/etc/config/app.conf ./local-app.conf`
Copy file to pod	`kubectl cp ./local.conf <pod>:/etc/config/local.conf`
Port forward to pod	`kubectl port-forward pod/<name> 8080:80`
Port forward to service	`kubectl port-forward svc/<name> 8080:80`

Multi-container / distroless

kubectl logs <pod> -c sidecar
kubectl exec <pod> -c app -- /bin/sh
kubectl debug pod/mypod -it --image=busybox:1.36.1 --target=app
kubectl debug pod/mypod -it --image=ubuntu --copy-to=debug-pod --share-processes

Exam gotchas

Events section at the bottom of kubectl describe pod: pull errors, scheduling failures, probe failures
CrashLoopBackOff: --previous to see what killed the container
kubectl top needs Metrics Server; if "Metrics API not available" → server not installed
Running ≠ healthy: check readiness + logs
Multi-container: specify -c

Practice scenario: Troubleshooting

Broken state: frontend 0 ready; backend empty Endpoints; CronJob no Jobs

Solution

Workflow:

get po → describe → logs / logs -p
Empty Endpoints → describe svc selector vs get po --show-labels
CronJob → describe cronjob (schedule, suspend, concurrency)

Typical fixes:

kubectl set image deployment/frontend nginx=nginx:1.25.3
# Fix Service selector to match Pod labels
# Fix CronJob schedule or re-apply manifest