NFS Server Provisioner deploys a NFS server to the Kubernetes cluster, exposing shared-storage via the NFS volume plugin.  Unlike the VsphereVolume plugin, the NFS plugin supports ReadWriteMany (RMX) access-modes, regardless of whether the pods are colocated on a common node.  This allows users to scale out web front-ends while maintaining content consistancy from a central, persistant data source.

$ kubectl config set-context $(kubectl config current-context) --namespace=kube-system

$ kubectl apply -f default-sc.yaml

https://docs.helm.sh/using_helm/#installing-helm or

$ curl https://raw.githubusercontent.com/kubernetes/helm/master/scripts/get > get_helm.sh

$ chmod 700 get_helm.sh

$ ./get_helm.sh

$ kubectl apply -f tiller-rbac.yaml

$ helm init --service-account tiller

Edited defaults to add persistance with a vsphere-volume, defined in the default storage class.

Note: Review sizes for both values.yml and corpweb-nfs-pvc.yaml before deploying.

Current values.yml - 120Gi

Current corpweb-nfs-pvc.yaml - 10Gi

$ helm install stable/nfs-server-provisioner --name k8s-nfs-server -f values.yaml

$ kubectl create ns corp-lab

$ kubectl config set-context $(kubectl config current-context) --namespace=corp-lab

$ kubectl apply -f corpweb-nfs-pvc.yaml

$ kubectl apply -f corpweb-app.yaml

$ kubectl get pods

$ kubectl cp html <podname>:/usr/share/nginx/

$ kubectl exec -it <podname> /bin/bash

$ ls /usr/share/nginx/html/

$ exit

$ kubectl scale --replicas=3 deploy/corpweb-deploy

$ kubectl get pods

Pick one of the new pods to verify the webcontent is available to it through the mounted NFS share

$ kubectl exec -it <2nd podname> /bin/bash

$ ls /usr/share/nginx/html/

$ exit

$ kubectl get ingress

Example: http://corpweb.k8s01-app.lab.local

$ kubectl cp htmlv2/vmware_luv.jpg <2nd podname>:/usr/share/nginx/html

Refresh the browser to view the updated image