Troubleshooting

A developer has a pod that won't deploy/update/work, how would you troubleshoot it?

• Run kubectl get pods -n $namespace to investigate the error to determine if there's an error message
• Make sure the pod is running, there isn't a CrashLoopBackoff, ImagePullErr, etc

How do you get all your running pods?

• kubectl get pods -A

How to tell why a container is failing?

• kubectl describe pod $pod_name, investigate the error or events
• kubectl logs $pod

How do you check the logs of a specific container in a pod?

• Use kubectl logs $pod_name -c $container_name to view logs for a specific container in a multi-container pod.

How do you troubleshoot a service that is not reachable?

• Run kubectl get svc to ensure the service is created and has a valid ClusterIP or external IP.
• Use kubectl describe svc $service_name to check for configuration issues.
• Verify that the pods backing the service are running and reachable.
• Verify that the pods and services are using the proper labels/selectors

How do you debug a pod stuck in ImagePullBackOff?

• Run kubectl describe pod $pod_name to verify the image name and tag are correct.
• Ensure the image is accessible from the container registry and that proper credentials are configured if needed.

How do you check resource usage for a pod?

• Use kubectl top pod $pod_name to view CPU and memory usage.
• Ensure the pod is not exceeding its resource limits defined in the deployment or pod specification.

How do you troubleshoot DNS issues in a pod?

• Use kubectl exec -it $pod_name -- nslookup $service_name to test DNS resolution.
• Verify that the kube-dns or CoreDNS pod is running and healthy using kubectl get pods -n kube-system.