kubernetes-troubleshoot

Troubleshoot CrashLoopBackOff by inspecting pod events, retrieving container logs, and describing the pod resource to identify application crashes or configuration errors. Check for mismatched image versions, missing environment variables, or resource constraints causing repeated failures.

Diagnose ImagePullBackOff by describing the pod to view pull error details, checking image registry credentials and availability, and reviewing node events. Verify the image URI is correct, the registry is accessible, and pull secrets are properly configured in your namespace.

Retrieve pod logs by using kubectl or MCP tools to fetch container output for debugging. Specify the namespace, pod name, and optionally the container name to access current or previous logs from failed containers.

Yes, you can scope resource inspection across namespaces and contexts to diagnose issues in multi-cluster environments. List and describe pods, deployments, nodes, and events while switching between namespaces and cluster contexts for complete visibility.

Debug pending pods by listing pod status, inspecting pod events for scheduling constraints, and describing node resources. Check for insufficient CPU/memory, taints, or affinity rules preventing pod placement on available nodes.

Yes, the troubleshooting workflow supports single and multi-cluster environments through context and namespace scoping. Guide stepwise debugging across multiple clusters by switching contexts and targeting specific namespaces for pod and workload inspection.