Kubernetes Pod logs
Kubectl logs <<Pod Name>>
To find the Kubernetes cluster events
Kubectl events
Docker logs
docker logs <<container name>>
To find the events occurred in docker daemon
docker events -- since <<date>>
Monitoring
Use Metric server implementation to monitor how much memory and CPU utilization for a particular pod or worker node.
There are other 3rd party tools are also available for monitoring ( Prometheus and Garfana)
Debugging and Troubleshooting
1. If the Pod is not getting created successfully then run kubectl describe pod <<podname>> command to check the events which are occured to created the pod and also check the properties of the pod in case you are getting unexpected value of any attribute
2. If you want to remove a pod under Replication controller or Deployment or a Service then run kubectl get pods -o wide command and identify the pod which is required to be deleted and delete with the help of kubectl delete pod <<pod name>> command
3. If a Pod is in running state but not working as per expected result, In that case check the log entries of the Pod
4. If a Pod is in the Pending state for a long time, In that case, check the worker node statuses by using kubectl get nodes command and check all nodes are in Ready State or not
5. If a Pod is not getting created on a particular node. In that case, check the node is tainted or not.
COMMENTS