kubernetes daemonset pod terminating for a time after being killed

Question

I have a Kubernetes Daemonset running on my GKE cluster version:

Server Version: version.Info{Major:"1", Minor:"9+", GitVersion:"v1.9.6-gke.1", GitCommit:"cb151369f60073317da686a6ce7de36abe2bda8d", GitTreeState:"clean", BuildDate:"2018-04-07T22:06:59Z", GoVersion:"go1.9.3b4", Compiler:"gc", Platform:"linux/amd64"}

When a pod gets deleted, my containers are handling the issued SIGTERM normally and shutting down. Both containers in the pod report the following immediately:

State: Terminated Reason: Completed Exit Code: 0 Started: Tue, 22 May 2018 17:09:39 -0500 Finished: Tue, 22 May 2018 17:24:11 -0500 Ready: False

However, the pod remains in Ready: 0/2, Status: Terminating for about 30 seconds afterward. Any ideas about why that might happen or how to diagnose it?

Answer 1

The default grace period is 30s. As any default, it may or may not fit your use case. There are two ways to change the grace period:

1.In the yaml file

For example, the following .yaml shows a simple DaemonSet config with a 30 seconds termination grace period

apiVersion: extensions/v1beta1
kind: DaemonSet
metadata:
.
.
.

terminationGracePeriodSeconds: 30

2.On the command line, you run “kubeclt delete”

For example,

kubectl delete pod NAME --grace-period=0

for a force delete as per @Zied post.

Answer 2

If we focus on deletion part you describe, we can refer to kubectl delete help to figure that pod ressource has grace-period

Some resources, such as pods, support graceful deletion. These resources > define a default period before they are forcibly terminated (the grace > period) but you may override that value with the --grace-period flag, or > pass --now to set a grace-period of 1.

You could try :

kubectl delete pod NAME --grace-period=0