Guide to Kubernetes Troubleshooting: Commands for Pods, Services, and Controllers

Table of Contents

1. Identifying Issues
2. Common Pod Commands
3. Common Controller Commands
4. Common Service Commands
5. Frequently Asked Questions (FAQs)

Kubernetes is a powerful container orchestration platform, but diagnosing and resolving issues can sometimes be daunting. This guide provides a comprehensive list of commands to troubleshoot problems with Pods, Controllers, and Services effectively.

---

1. Identifying Issues

Before troubleshooting, determine the type of issue:

• Is it related to a Pod (e.g., stuck in Pending, crashing)?
• Is it tied to a Controller like a Deployment or StatefulSet?
• Or is it a problem with a Service, such as inaccessible endpoints?

Based on the type of issue, use the respective commands in the following sections.

---

2. Common Pod Commands

Q: How do I troubleshoot a Pod stuck in the Pending state?

• Answer: Check for resource availability and node status using: kubectl describe pod ${POD_NAME} Look for events like "Insufficient CPU" or "NodeAffinity violations."

Q: What if a Pod crashes repeatedly?

• Answer: Use the logs command to identify errors: kubectl logs ${POD_NAME} ${CONTAINER_NAME}

Additional commands for Pods include:

• Get Pod Status: kubectl -n ${NAMESPACE} get pod -o wide
• Access the Pod for Debugging: kubectl exec -it ${POD_NAME} /bin/bash

---

3. Common Controller Commands

Controllers like Deployments and StatefulSets manage the lifecycle of Pods.

Q: What should I do if my Pods fail to start after a Deployment update?

• Answer: Use: kubectl describe deployment ${DEPLOYMENT_NAME} Check for events related to image pull errors, configuration mismatches, or insufficient resources.

Other useful commands include:

• Check Deployment Status: kubectl get deployment -o wide
• View Deployment Configuration: kubectl get deployment ${DEPLOYMENT_NAME} -o yaml

---

4. Common Service Commands

Q: How do I diagnose a Service with no endpoints?

• Answer: Run: kubectl describe svc ${SERVICE_NAME} Verify the Endpoints field. If empty, ensure the Service's selector matches the Pod labels.

Additional commands include:

• View Service Status: kubectl get svc -o wide
• Check Load Balancer and Port Information: kubectl describe svc ${SERVICE_NAME}

---

5. Frequently Asked Questions (FAQs)

Q: How can I monitor resource usage for Pods?

• Answer: Use the following command to get CPU and memory usage: kubectl top pod --namespace=${NAMESPACE}

Q: What’s the best way to debug network issues in Kubernetes?

• Answer: Check Pod connectivity using kubectl exec with tools like curl or ping. kubectl exec -it ${POD_NAME} -- curl ${SERVICE_NAME}

Q: Why is my Deployment not scaling?

• Answer: Inspect the HPA (Horizontal Pod Autoscaler) settings using: kubectl get hpa -n ${NAMESPACE}

---

Conclusion

With these commands and insights, Kubernetes administrators can efficiently identify and resolve cluster issues, ensuring smoother application operations. Regular monitoring and debugging practices can help minimize downtime and optimize performance.