• Create the pod with kubectl create -f nginx.yaml.
  • Why is a pod in a Pending state?
    • The Scheduler is what schedules the pod.
    • Run kubectl describe pod nginx
      • Check the Node field and it will be set to <none>
        • The answer then is No Scheduler Present
  • Use kubectl replace --force -f nginx.yaml forcefully deletes and recreates the pod, no need to use the kubectl delete command.
  • To continually monitor pods, use kubectl get pods --watch
  • Cannot move one running pod from one node to another (similar to a running process on bare metal).
    • Can only delete pods on one system or node and recreate it on another.
  • When you delete a pod, in the backend it has to terminate the nginx process.
    • It sends a kill signal to the process and it will take time to process the request and then gracefully shutdown.
  • Use the wide option to see which node the pod is scheduled on:
    kubectl get pods -o wide
    

Updated: