Kpanda Pod Diagnosis

Diagnose unhealthy or failing pods through a standardized inspection workflow.

REQUIRED SUB-SKILL: Use dce for all command execution, auth checks, and catalog discovery.

Workflow

Step 1 — Identify Target Pods

Run dce container-management core list-cluster-pods --cluster <cluster> -o json to list all pods. Filter for non-Running/Completed phases:

• Pending — scheduling or resource issues
• Failed — terminal crash
• Unknown — control plane communication failure
• Evicted — preemption or node pressure
• CrashLoopBackOff / OOMKilled — visible in container states