Troubleshooting K8s

How do I troubleshoot Kubernetes deployment errors?

only 3 cmd:
for troubleshooting:
kubectl describe,log,events

Troubleshooting Kubernetes deployments requires a systematic approach to identify why resources are failing to reach a healthy state. According to the video, here are the most effective methods to debug your cluster:

• Examine Deployment Status: Use kubectl get pods and kubectl describe pod <pod-name> to identify status messages like CrashLoopBackOff, ImagePullBackOff, or Pending. Describing the pod helps reveal specific errors from the scheduler or container runtime (1:17:00).
• Check Container Logs: If a pod is crashing, inspect the logs of specific containers using kubectl logs <pod-name> -c <container-name>. This is crucial for multi-container pods to pinpoint which part of the application is failing (1:18:00 - 1:23:00).
• Monitor Events: Use kubectl get events to see the history of cluster activities. This often reveals issues with StorageClasses, missing permissions, or resource constraints (2:10:00).
• Analyze Configuration: If you encounter errors related to StatefulSets or persistent storage (like PVCs staying in a Pending state), verify your StorageClass settings and ensure the configuration matches your cloud environment's requirements, such as using GP3 storage classes on AWS (2:11:00 - 2:16:00).
• Verify Infrastructure Dependencies: Ensure that networking, security groups, and IAM roles (if using IRSA) are correctly configured, as infrastructure-level issues often mask themselves as application errors (3:24:00 - 3:28:00).