Reading kubectl describe Output for Failures

The kubectl describe command provides a comprehensive view of a resource's current state, conditions, and recent events. Learning to read its output is one of the most important debugging skills in Kubernetes.

Key Sections in describe Output

kubectl describe pod my-pod

Focus on these sections when troubleshooting:

• Status: The current phase (Pending, Running, Failed, Succeeded)
• Conditions: Boolean statuses like PodScheduled, Initialized, ContainersReady, Ready
• State / Last State: Current and previous container state with exit codes and reasons
• Events: Recent events attached to this specific resource

Debugging Pending Pods

A pod stuck in Pending has not been scheduled to a node. Look at Conditions and Events:

kubectl describe pod pending-pod

Common causes include:

• Insufficient resources: 0/3 nodes are available: 3 Insufficient cpu
• No matching nodes: Taints, tolerations, or nodeSelector mismatches
• Unbound PVC: persistentvolumeclaim "data" not found

Debugging ImagePullBackOff

The container image could not be pulled. Check the Events section:

kubectl describe pod bad-image-pod
# Events will show:
#   Failed to pull image "myapp:wrong-tag": manifest unknown

Common causes: wrong image name or tag, private registry without imagePullSecrets, network issues.

Debugging CrashLoopBackOff

The container starts but exits immediately. The describe output shows:

kubectl describe pod crashing-pod
# Last State:  Terminated
#   Exit Code: 1
#   Reason:    Error

Pair this with kubectl logs --previous to see what the application printed before crashing.

Common Failure Pattern Summary

Always start with describe to get the high-level picture, then drill into logs and events for details.