Exercise

  1. Create a Pod named debug with one container based on alpine:3.15 executing the command sleep 10000. Make sure the Pod is running fine

  2. Measure the time needed to delete the Pod. How to do explain that ?

  3. Which command could be used to delete the Pod right away ?

Solution
  1. Create a Pod named debug with one container based on alpine:3.15 executing the command sleep 10000. Make sure the Pod is running fine

Creation of the Pod:

k run debug --image=alpine:3.15 --command sleep 10000

Making sure the Pod is running:

k get po debug
NAME    READY   STATUS    RESTARTS   AGE
debug   1/1     Running   0          4s
  1. Measure the time needed to delete the Pod. How to do explain that ?

It takes around 30 seconds for the pod to be deleted:

time k delete po debug
pod "debug" deleted

real	0m30.682s  <- you've been waiting more than 30 seconds for the pod to the deleted
user	0m0.096s
sys	  0m0.037s

There are some cases where the sigterm signal is not forwarded to the container. In that case, a sigkill is sent after 30 seconds.

  1. Which command could be used to delete the Pod right away ?

You can force the Pod deletion with the following options:

k delete po debug --force --grace-period=0