Patch k8s instructons

gistfile1.txt

# Steps to patch k8s 1.11 with umount timeout fix

Summary:
---
* Custom image id: husseingalal/hyperkube-rancher:v1.11.6
* As Rancher 2.1.x doesn't support passing an arbitrary kubernetes version to the cluster create/update, the custom image has to be pre-downloaded to the hosts and tagged as the image officially supported by Rancher 2.1.3.
* Rancher engineer verified the fix provided with the patch, and confirms that reducing the hardcoded volume unmount timeout, results in volume getting unbound from the shutdown host in 10 seconds as opposed to 6 minutes seen on non-patched k8s version.

Instructions:
---
1- Run k8s version `1.10.11-rancher1-1` on rancher `2.1.3` with vsphere cloud provider.
2- After the installation is finished, connect on each node and run the following commands to download and retag the custom image:
```
$ docker pull husseingalal/hyperkube-rancher:v1.11.6
$ docker tag husseingalal/hyperkube-rancher:v1.11.6 rancher/hyperkube:v1.11.5-rancher1
```
3- Update the kubernetes version to `v1.11.5-rancher1-1` which will upgrade the cluster to the customized k8s image.
4- to make sure you have the customized image, run the following command:
```
$ kubectl get nodes
NAME                                          STATUS    ROLES                      AGE       VERSION
ip-172-31-16-35.us-west-2.compute.internal    Ready     controlplane,etcd,worker   10m       v1.11.6-1+23b9c87990eee9
ip-172-31-19-250.us-west-2.compute.internal   Ready     controlplane,etcd,worker   10m       v1.11.6-1+23b9c87990eee9
ip-172-31-26-212.us-west-2.compute.internal   Ready     controlplane,etcd,worker   10m       v1.11.6-1+23b9c87990eee9
```
The version should say v1.11.6