Setting Up NFS Storage in Kubernetes with Helm

The article demonstrates how to provision NFS storage in a Kubernetes environment, starting with verification of cluster component health using kubectl get cs and noting the deprecation warning for ComponentStatus.

It then creates the NFS directory on the host with mkdir /data/nfs, installs the NFS utilities via yum -y install nfs-utils rpcbind, and configures the export file /etc/exports to share the directory: /data/nfs *(rw,sync,no_root_squash,no_subtree_check) After starting the NFS and rpcbind services using systemctl start nfs && systemctl start rpcbind, a dedicated namespace for NFS is created with kubectl create ns nfs.

Next, Helm is installed by extracting the binary from helm-v3.4.2-linux-amd64.tar.gz, moving it to /usr/bin/helm, and verifying the installation with helm version. Helm repositories are added ( helm repo add helm-stable https://charts.helm.sh/stable and

helm repo add aliyun https://kubernetes.oss-cn-hangzhou.aliyuncs.com/charts

) and updated.

The NFS client provisioner chart is searched ( helm search repo helm-stable/nfs-client-provisioner) and a values file nfs.yaml is prepared with the NFS server address, path, storage class name, reclaim policy, and image details.

Finally, the provisioner is installed with:

helm install nfs-storage -n nfs --values nfs.yaml helm-stable/nfs-client-provisioner --version 1.2.11

The deployment can be verified by listing pods in the nfs namespace using kubectl get pods -n nfs, confirming that the nfs-client-provisioner pod is running.