luc.run

Exercise

Run a Deployment named www with 6 replicas of Pod based on nginx:1.20
Where are the Pods running ?
Using a single command, make sure the Pods are evicted from worker2
Delete the Deployment and revert your changes back

Documentation

https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/

Solution

Run a Deployment with 6 replicas of Pod based on nginx:1.20

k create deploy www --image=nginx:1.20 --replicas=6

Where are the Pods running ?

Pods are running on worker1 and worker2 nodes (as master node as a NoSchedule taint preventing application pods from been scheduled on that node).

k get po -o wide
NAME                   READY   STATUS    RESTARTS   AGE   IP          NODE      NOMINATED NODE   READINESS GATES
www-644dfdf68b-2b42t   1/1     Running   0          3s    10.32.0.3   worker1   <none>           <none>
www-644dfdf68b-6vxmz   1/1     Running   0          3s    10.38.0.2   worker2   <none>           <none>
www-644dfdf68b-jrrd7   1/1     Running   0          3s    10.38.0.1   worker2   <none>           <none>
www-644dfdf68b-kgrc6   1/1     Running   0          3s    10.38.0.5   worker2   <none>           <none>
www-644dfdf68b-lx2sp   1/1     Running   0          3s    10.32.0.5   worker1   <none>           <none>
www-644dfdf68b-mvsfv   1/1     Running   0          3s    10.32.0.4   worker1   <none>           <none>

Using a single command, make sure the Pods are evicted from worker2

Adding a taint with the NoExecute effect on a node will evict all the Pods that do not tolerate that taint

k taint node worker2 app=blue:NoExecute

The Pods now all run on worker1

k get po -o wide
www-644dfdf68b-2b42t   1/1     Running   0          98s   10.32.0.3    worker1   <none>           <none>
www-644dfdf68b-5wmfs   1/1     Running   0          15s   10.32.0.8    worker1   <none>           <none>
www-644dfdf68b-fbf2v   1/1     Running   0          15s   10.32.0.10   worker1   <none>           <none>
www-644dfdf68b-lx2sp   1/1     Running   0          98s   10.32.0.5    worker1   <none>           <none>
www-644dfdf68b-mvsfv   1/1     Running   0          98s   10.32.0.4    worker1   <none>           <none>
www-644dfdf68b-p8xg2   1/1     Running   0          15s   10.32.0.9    worker1   <none>           <none>

Delete the Deployment and revert your changes back

Deletion of the Deployment

k delete deploy www

Remove the NoExecute taint from worker2

k taint node worker2 app-