Draining nodes

In this training, we will learn about how to drain and uncordon a node.

Navigate to the folder 19_nodedrain from CLI, before you get started.

Create the deployment, pod and daemonset

kubectl create -f .

Get the node
```
kubectl get pods -o wide
```
Drain the node, where the pod my-pod is running
```
kubectl drain <NODE-NAME>
```
This will get aborted due to my-daemonset and my-pod

Try again

kubectl drain <NODE-NAME> --ignore-daemonsets --force

Take a look at the pods on that node
```
kubectl get pods -o wide
```
Note that the pod my-pod has gone forever. The pods for the deployment my-deployment got rescheduled to the other node and the pods for the daemonset my-daemonset are untouched.

kubectl uncordon <NODE-NAME>

Note that the pod my-pod really has gone forever. The uncordon operation does not cause any re-scheduling happening.

kubectl delete deployment my-deployment
kubectl delete daemonset my-daemonset