DOE306 - Volume management in K8s

• DOE306 - Volume management in K8s
  • Contents
  • Overview
    • Volumes
    • Persistent Volumes
    • Volumes types
  • LAB #1 - Using K8s Volumes
    • 1.1 - Volumes and volumeMounts
    • 1.2 - Sharing volumes between containers
  • LAB #2 - Persistent Volumes
    • 2.1 - Storage Classes
    • 2.2 - Persistent Volumes
    • 2.3 - Persistent Volume Claims
    • 2.4 - Using a PersistentVolumeClaim in a pod
    • 2.5 - Resizing a PersistentVolumeClaim

The file system of a container in a pod is ephemeral, i.e. it exists only during the life cycle of the container. If the container is deleted or re-created, the file system is lost.

Volumes enable data to be stored outside the container's file system, while still allowing the container to access it.

A Persistent Volume is an abstract resource that can be consumed by pods. To access the Persistent Volume, the pod needs a Persistent Volume Claim to mount the Persistent Volume in the pod.

Volumes and Persistent Volumes have a Volume Type (Type of Volume). The Volume Type determines the Storage Method of the data. Storage methods include:

• NFS,
• AWS,
• Azure,
• GCP,
• ConfigMaps,
• Secrets,

Volumes are configured in the pod specification, not the container. The two most important Volume Types are hostPath and emptyDir:

• hostPath,
  • Data is stored locally in a static directory on the K8s node,
• emptyDir,
  • Data is stored locally in a dynamic directory,
  • The directory exists only while the pod exists on the node,
  • K8s deletes the directory and data when the pod is deleted or moved,
  • This Volume Type is mainly used to share data between two containers in a pod.

A volumeMount is configured in the container specification, not the pod.

Start by creating the volume.yaml file:

root@kubemaster:~# vi volume.yaml
root@kubemaster:~# cat volume.yaml
apiVersion: v1
kind: Pod
metadata:
  name: volumepod
spec:
  restartPolicy: Never
  containers:
  - name: busybox
    image: busybox
    command: ['sh', '-c', 'echo Success! > /output/success.txt']
    volumeMounts:
    - name: myvolume
      mountPath: /output
  volumes:
  - name: myvolume
    hostPath:
      path: /var/data

Create the volumepod pod:

root@kubemaster:~# kubectl create -f volume.yaml
pod/volumepod created

Identify the node on which the pod runs:

root@kubemaster:~# kubectl get pod volumepod -o wide
NAME        READY   STATUS      RESTARTS   AGE     IP               NODE                       NOMINATED NODE   READINESS GATES
volumepod   0/1     Completed   0          3m10s   192.168.150.41   kubenode2.ittraining.loc   <none>           <none>

Connect to the identified node:

root@kubemaster:~# ssh -l trainee kubenode2
trainee@kubenode2's password: trainee
Linux kubenode2.ittraining.loc 4.9.0-19-amd64 #1 SMP Debian 4.9.320-2 (2022-06-30) x86_64

The programs included with the Debian GNU/Linux system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
permitted by applicable law.
Last login: Sun Sep 4 13:24:39 2022 from 192.168.56.2

Check for the presence and content of the file /var/data/success.txt:

trainee@kubenode2:~$ cat /var/data/success.txt
Success!

Return to the kubemaster VM:

trainee@kubenode2:~$ exit
logout
Connection to kubenode2 closed.
root@kubemaster:~# 

Now create the shared.yaml file:

root@kubemaster:~# vi shared.yaml
root@kubemaster:~# cat shared.yaml
apiVersion: v1
kind: Pod
metadata:
  name: sharedvolume
spec:
  containers:
  - name: busybox1
    image: busybox
    command: ['sh', '-c', 'while true; do echo Success! > /output/output.txt; sleep 5; done']
    volumeMounts:
      - name: myvolume
        mountPath: /output
  - name: busybox2
    image: busybox
    command: ['sh', '-c', 'while true; do cat /input/output.txt; sleep 5; done']
    volumeMounts:
    - name: myvolume
      mountPath: /input
  volumes:
  - name: myvolume
    emptyDir: {}

Create the two pods:

root@kubemaster:~# kubectl create -f shared.yaml
pod/sharedvolume created

Check that both pods are running:

root@kubemaster:~# kubectl get pods sharedvolume
NAME           READY   STATUS    RESTARTS   AGE
sharedvolume   2/2     Running   0          5m55s

Now look at the logs for the second container:

root@kubemaster:~# kubectl logs sharedvolume -c busybox2
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!
Success!

• StorageClassName,
  • A StorageClassName is used to specify the StorageClass.
• StorageClass,
  • A StorageClass is used to specify the type of storage service used, e.g. local disk, cloud etc,
  • If the allowVolumeExpansion value is true and the storage service type allows it, a PersistentVolumeClaim can be hot-resized.

Create the localdisk.yaml file to define the StorageClass called localdisk:

root@kubemaster:~# vi localdisk.yaml
root@kubemaster:~# cat localdisk.yaml 
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: localdisk
provisioner: kubernetes.io/no-provisioner
allowVolumeExpansion: true

Create the StorageClass localdisk :

root@kubemaster:~# kubectl create -f localdisk.yaml

Create the mypv.yaml file to define the PersistentVolume called mypv:

root@kubemaster:~# vi mypv.yaml
root@kubemaster:~# cat mypv.yaml 
kind: PersistentVolume
apiVersion: v1
metadata:
  name: mypv
spec:
  storageClassName: localdisk
  persistentVolumeReclaimPolicy: Recycle
  capacity:
    storage: 1Gi
  accessModes:
    - ReadWriteOnce
  hostPath:
    path: /var/output

There are four types of accessModes:

• ReadWriteOnce or RWO,
  • the volume can only be mounted by a single node,
• ReadOnlyMany or ROX,
  • the volume can be mounted read-only by several nodes,
• ReadWriteMany or RWX,
  • the volume can be mounted as read-write by several nodes,
• ReadWriteOncePod or RWOP,
  • the volume can only be mounted by a single pod.

There are three types of PersistentVolumeReclaimPolicy:

• Retain,
  • Data is not deleted when a PersistentVolumeClaim is deleted,
• Delete,
  • Automatically deletes the storage resource when a PersistentVolumeClaim is deleted. Note that Delete only works with public cloud services such as AWS, GCP etc,
• Recycle,
  • Automatically deletes data. Note that Recycle allows immediate reuse of storage resources freed up when a PersistentVolumeClaim is deleted.

Create the PersistentVolume mypv:

root@kubemaster:~# kubectl create -f mypv.yaml
persistentvolume/mypv created

Check the PersistentVolume status:

root@kubemaster:~# kubectl get pv
NAME   CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS      CLAIM   STORAGECLASS   REASON   AGE
mypv   1Gi        RWO            Recycle          Available           localdisk               5m28s

• A PersistentVolumeClaim represents a user's request for a storage resource,
  • The PersistentVolumeClaim specifies a StorageClassName, an AccessMode and a size,
• When created, the PersistentVolumeClaim searches for a PersistentVolume capable of satisfying the request,
  • If the result of the search is positive, the PersistentVolumeClaim is automatically linked to the PersistentVolume,
  • Otherwise, the PersistentVolumeClaim remains on hold until a PersistentVolume capable of satisfying the request is created,
  • To resize a PersistentVolumeClaim without interrupting service, change the value of spec.resources.requests.storage in the yaml file.

Create the mypvc.yaml file to define the PersistentVolumeClaim called my-pvc:

root@kubemaster:~# vi mypvc.yaml
root@kubemaster:~# cat mypvc.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: my-pvc
spec:
  storageClassName: localdisk
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 100Mi

Create the PersistentVolumeClaim my-pvc:

root@kubemaster:~# kubectl create -f mypvc.yaml
persistentvolumeclaim/my-pvc created

Check the PersistentVolume status:

root@kubemaster:~# kubectl get pv
NAME   CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM            STORAGECLASS   REASON   AGE
mypv   1Gi        RWO            Recycle          Bound    default/my-pvc   localdisk               9m33s

Check the PersistentVolumeClaim status:

root@kubemaster:~# kubectl get pvc
NAME     STATUS   VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE
my-pvc   Bound    mypv     1Gi        RWO            localdisk      72s

Create the mypvcpod.yaml file to define the pod called pv-pod:

root@kubemaster:~# vi mypvcpod.yaml
root@kubemaster:~# cat mypvcpod.yaml
apiVersion: v1
kind: Pod
metadata:
  name: pv-pod
spec:
  restartPolicy: Never
  containers:
  - name: busybox
    image: busybox
    command: ['sh', '-c', 'echo Success! > /output/success.txt']
    volumeMounts:
    - name: pv-storage
      mountPath: /output
  volumes:
  - name: pv-storage
    persistentVolumeClaim:
      claimName: my-pvc

Create the pv-pod pod:

root@kubemaster:~# kubectl create -f mypvcpod.yaml
pod/pv-pod created

Modify the storage: value of the PersistentVolumeClaim:

root@kubemaster:~# kubectl edit pvc my-pvc --record
...
spec:
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 200Mi
  storageClassName: localdisk
  volumeMode: Filesystem
  volumeName: mypv
...

Save the modification:

root@kubemaster:~# kubectl edit pvc my-pvc --record
Flag --record has been deprecated, --record will be removed in the future
persistentvolumeclaim/my-pvc edited

Delete the pv-pod pod and the my-pvc PersistentVolumeClaim:

root@kubemaster:~# kubectl delete pod pv-pod
pod “pv-pod” deleted

root@kubemaster:~# kubectl delete pvc my-pvc
persistentvolumeclaim “my-pvc” deleted

Check the PersistentVolume status:

root@kubemaster:~# kubectl get pv
NAME   CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS      CLAIM   STORAGECLASS   REASON   AGE
mypv   1Gi        RWO            Recycle          Available           localdisk               23m

Copyright © 2025 Hugh Norris