Container Service for Kubernetes:Configure cloud disk persistent storage for StatefulSets

The following is a sample volumeClaimTemplates configuration:

apiVersion: apps/v1
kind: StatefulSet
# ...
spec:
  # ...
  volumeClaimTemplates:
  - metadata:
      name: data-volume                        # Name of the PVC template
    spec:
      accessModes: [ "ReadWriteOnce" ]
      storageClassName: "alicloud-disk-essd"   # Specify the storage type
      resources:
        requests:
          storage: 20Gi                        # The requested storage capacity

How it works

Step 1: Deploy a StatefulSet with persistent storage

This example deploys a Service and a StatefulSet with two replicas. The StatefulSet uses volumeClaimTemplates to automatically create a 20 GiB cloud disk for each replica.

1. Create a file named statefulset.yaml.

   The following table describes the parameters in volumeClaimTemplates:

   Parameter	Description
   accessModes	The access mode of the volume. ReadWriteOnce means the volume can be mounted as read-write by a single node at a time.
   storageClassName	The name of the StorageClass to use. alicloud-disk-essd is a default StorageClass provided by Container Service for Kubernetes (ACK) for creating enterprise SSDs (ESSDs) with a default performance level (PL) of PL1. These disks use pay-as-you-go billing. For more information, see Billing of block storage and Prices of block storage.
   storage	The capacity of the disk volume.

   YAML template

   apiVersion: v1
   kind: Service
   metadata:
     name: nginx
     labels:
       app: nginx
   spec:
     ports:
     - port: 80
       name: web
     # Set clusterIP to "None" to indicate a Headless Service
     clusterIP: None
     selector:
       app: nginx
   ---
   apiVersion: apps/v1
   kind: StatefulSet
   metadata:
     name: web
   spec:
     selector:
       matchLabels:
         app: nginx
     # The serviceName must match the name of the Headless Service defined above
     serviceName: "nginx"
     replicas: 2
     template:
       metadata:
         labels:
           app: nginx
       spec:
         containers:
         - name: nginx
           image: anolis-registry.cn-zhangjiakou.cr.aliyuncs.com/openanolis/nginx:1.14.1-8.6
           ports:
           - containerPort: 80
             name: web
           # Mount the PV to a specified path in the container
           volumeMounts:
           - name: disk-essd
             mountPath: /data
     # PVC template used by the StatefulSet to create a PVC for each pod
     volumeClaimTemplates:
     - metadata:
         name: disk-essd
       spec:
         # Define the volume access mode
         accessModes: [ "ReadWriteOnce" ]
         # Specify the StorageClass for dynamic PV provisioning
         storageClassName: "alicloud-disk-essd"
         resources:
           requests:
             # Define the requested storage capacity for each PVC
             storage: 20Gi

2. Deploy the StatefulSet.

   kubectl create -f statefulset.yaml

3. Verify that pods are running.

   kubectl get pod -l app=nginx

4. View the PVCs to confirm that the system has automatically created and bound a corresponding PVC for each pod.

   kubectl get pvc

   Expected output:

   NAME              STATUS   VOLUME                   CAPACITY   ACCESS MODES   STORAGECLASS         VOLUMEATTRIBUTESCLASS   AGE
   disk-essd-web-0   Bound    d-m5eb5ozeseslnz7zq54b   20Gi       RWO            alicloud-disk-essd   <unset>                 3m31s
   disk-essd-web-1   Bound    d-m5ecrvjrhqwehgzqpk5i   20Gi       RWO            alicloud-disk-essd   <unset>                 48s

Step 2: Validate the storage lifecycle

Observe the creation, retention, and reuse of associated PVCs by scaling out, scaling in, then scaling out again.

Step 3: Validate data persistence after a pod failure

Verify that data stored on the disk persists after a pod is recreated by writing data, deleting the pod, then checking for the data.

1. Write test data to the pod.

   Using pod web-1 as an example, create a test file in the mounted disk path /data.

   kubectl exec web-1 -- touch /data/test
   kubectl exec web-1 -- ls /data

   Expected output:

   lost+found
   test

2. Simulate a pod failure by deleting the pod.

   kubectl delete pod web-1

   Run kubectl get pod -l app=nginx again, you will see that a new pod named web-1 is automatically created.

3. Verify the data in the new pod.

   Check the /data directory in the new web-1 pod.

   kubectl exec web-1 -- ls /data

   The test file you created still exists. This confirms that data persists even if the pod is deleted and recreated.

   lost+found
   test

Application in production

• Cost and resource management: When you scale in or delete a StatefulSet, the associated PVCs and disks are retained by default. These retained resources continue to incur fees. Be sure to manually clean up any unused PVCs and PVs to avoid unnecessary charges.

• Data security and backup: Persistent storage ensures high availability during pod failures, but it is not a data backup solution. For critical data, use the backup center to perform regular backups.

• High availability and disaster recovery: Disks are zonal resources and cannot be mounted across zones. For cross-zone disaster recovery, use a disk type that supports cross-zone data replication, such as regional ESSDs.

References

See Disk volume FAQ for troubleshooting.