Objective
This guide documents how to manage persistent volumes on a Tanzu Kubernetes Grid cluster.
Requirements
- Being an administrative contact of your Hosted Private Cloud infrastructure to receive login credentials
- A user account with access to the OVHcloud Control Panel
- A user account with access to vSphere
- You need to have deployed the Workload TKG cluster using the Tanzu Management Cluster Grid Administration guide.
Introduction
Persistent volumes are used to store data permanently on a Kubernetes cluster. This mechanism is based on Storage Classes. There are various Storage Classes. Read the Kubernetes Storage Classes guide for more information.
When deploying a WorkLoad cluster, a Storage Class CSI provisioner is created and points to the Datastore folder that contains the virtual machines in the WorkLoad cluster.
From your VMware cluster, go to the inventory. Select the storage icon to the left, and go to the datastore where your WorkLoad cluster was deployed. Go to the Files tab and click the fcd folder.
The folder is empty because the WorkLoad cluster does not yet use persistent volumes.
You can create additional Storage Classes for each WorkLoad cluster.
Instructions
We will connect to a WorkLoad cluster from the console of the Bootstrap virtual machine. You can use the Tanzu Management Cluster Grid Administration guide to create and manage a workload cluster.
From the Bootstrap virtual machine console, run this command to view the contexts that can be used on this cluster:
# Display all the contexts of your TANZU KUBERNETES GRID cluster
kubectl config get-contexts
Run this command to use the WorkLoad cluster:
# Connect to WorkLoad cluster
kubectl config use-context tkgm-workload-cluster-admin@tkgm-workload-cluster
Displaying Existing Storage Classes
For information about the Storage Classes in a WorkLoad cluster, run these commands:
# Display Storage Classes
kubectl get storageclass
# Storage Class description
kubectl describe storageclass nomclasse
Creating a storage class on another datastore
On our VMware cluster, we have two datastores connected to NFS servers. One of the datastores contains the virtual machines in the WorkLoad cluster, as well as the fcd folder used by the Storage Class in the Workload cluster.
We will create a new Storage Class on the second datastore.
Go back to your VMware cluster in storage management, select the second datastore, and click Summary in the tab on the left.
Copy the URL below “Type:NFS3”.
Go to the console of the Bootstrap virtual machine, edit a new file named secondstorageclass.yaml with this content:
kind:StorageClassapiVersion:storage.k8s.io/v1metadata:name:secondstorageclassannotations:storageclass.kubernetes.io/is-default-class:"true"provisioner:csi.vsphere.vmware.comparameters:datastoreurl:"ds:///vmfs/volumes/xxxxxxxxxxxxxxxxx/"
Edit the file by replacingds:///vmfs/volumes/xxxxxxxxxxxxxxxxx/with the URL you just copied.
Next, run this command:
# Creating storageclass from yaml file
kubectl apply -f secondstorageclass.yaml
# Display storageclasses
kubectl get storageclass
We are now seeing two Storage Classes:
NAME PROVISIONER RECLAIMPOLICY VOLUMEBINDINGMODE ALLOWVOLUMEEXPANSION AGE
default (default) csi.vsphere.vmware.com Delete Immediate true 3d18h
secondstorageclass (default) csi.vsphere.vmware.com Delete Immediate false 12s
tanzu@bootstrap:~$
Creating a persistent volume in the default Storage class
Create a file named default-pvc-storage.yaml with this content:
kind:persistantVolumeClaimapiVersion:v1metadata:name:default-pvc-storagespec:accessModes:-ReadWriteOncestorageClassName:"default"resources:requests:storage:2Gi
The persistent storage name is next to name, storageClassName contains the name of the Storage Class that is used for this volume.
Run this command to create the persistent volume:
# Create a namespace that will be used for my persistent volume.
kubectl create namespace myspace
# Applying the configuration file to the created namespace.
kubectl apply -f default-pvc-storage.yaml -n myspace
# Display persistent volumes of created namespace
kubectl get pv,pvc -n myspace
Go back to the inventory in your vCenter interface, click on the DataCenter icon on the left, then go to the Monitor tab on the right, and click on Container Volumes to see the persistent volumes.
The persistent volume that has been created is displayed, and to the right you can see the name of the datastore on which it is stored.
Click the notebook icon to the left of the volume to view the details.
The information about this persistent storage is displayed, and corresponds to what was created using Kubernetes commands.
Go to the Datastore that is used by default, right-click the Files tab and scroll through the folders in the Datastore to the fcd folder.
You see that the folder contains two files, a vmdk file that contains the persistent volume data and an associated temporary file.
Creating a persistent volume on the second Storage Class
Return to the Bootstrap virtual machine and use the command line.
Create a file named second-storage-pvc.yaml:
kind:persistantVolumeClaimapiVersion:v1metadata:name:second-storage-pvcspec:accessModes:-ReadWriteOncestorageClassName:"secondstorageclass"resources:requests:storage:2Gi
The file uses the same syntax as the first persistent storage, but with a different name and Storage Class.
Run this command to create the persistent volume in the myspace namespace:
# Applying the configuration file to the created namespace.
kubectl apply -f second-storage-pvc.yaml -n myspace
# Display persistent volumes of created namespace
kubectl get pv,pvc -n myspace
The persistent volume is created on the second Datastore.
Return to the vCenter interface. You will see that you do not have any new files in the fcd folder.
Right-click the second Datastore, go to the fcd folder for that datastore. You will see that you have two new files, as in the first datastore.
Go back to the Datacenter at the root of the datacenters, click on the Monitor tab, and choose Container volumes to see the two persistent volumes appear with their locations in the datastores.