Help Center/ Cloud Container Engine/ User Guide/ Storage/ EVS/ Dynamically Mounting an EVS Disk to a StatefulSet

Updated on 2026-08-29 GMT+08:00

Dynamically Mounting an EVS Disk to a StatefulSet

Scenarios

Dynamic mounting is available only for creating a StatefulSet. It is implemented through a PVC template (the volumeClaimTemplates field) and depends on dynamic creation of PVs through StorageClass. In this mode, each pod in a StatefulSet is associated with a unique PVC and PV. After a pod is rescheduled, the original data can still be mounted to it based on the PVC name. In the common mounting mode for a Deployment, if ReadWriteMany is supported, multiple pods of the Deployment will be mounted to the same underlying storage.

When updating a StatefulSet in Kubernetes, it is not allowed to add or delete the volumeClaimTemplates field. This field can only be configured during the creation of the StatefulSet.

Working Principles

Workload creation or scale-out
When a StatefulSet is created or scaled out, a dedicated PVC is created for each pod (for example, web-0) based on the volumeClaimTemplates template. The PVC is named in the format: <template-name>-<StatefulSet-name>-<pod-sequence-number>. Parameters such as the storage class and resource request in the template are reused, and Kubernetes automatically binds the matched PV to the PVC. In this case, each pod has a PVC bound. The storage resources are configured when the pod is initialized, ensuring that each pod has independent and persistent storage.
Scale-in
When a StatefulSet is scaled in (for example, the StatefulSet is scaled from three pods to two pods), Kubernetes deletes only the pod with the largest sequence number (for example, web-2). The associated PVC and PV are not automatically deleted, which means that the storage resources are retained. Even if the pod is removed, the associated PVC still exists, and data is not lost. This provides a data basis for subsequent pod restoration.
Scale-out or fault recovery again
If a StatefulSet is scaled out again (for example, the StatefulSet is scaled from two pods to three instances) or the fault is rectified, the StatefulSet re-creates the web-2 pod and preferentially associates the retained PVC with the same name (for example, data-web-2). The original PV is directly mounted to the new pod. You do not need to reconfigure the storage, and data is fully reused. This process ensures that the expanded pods can still connect to the original storage resources, ensuring data consistency of stateful applications.

Prerequisites

You have created a cluster and installed the CCE Container Storage (Everest) add-on in the cluster.
To create a cluster using commands, ensure kubectl is used. For details, see Accessing a Cluster Using kubectl.

Dynamically Mounting an EVS Disk

CCE allows you to dynamically mount EVS volumes to workloads using the console or kubectl.

Log in to the CCE console and click the cluster name to access the cluster console.
Choose Workloads in the navigation pane. In the right pane, click the StatefulSets tab.
Click Create Workload in the upper right corner. On the displayed page, click Data Storage in the Container Information area under Container Settings and choose Add Volume > VCT.

Click Create PVC. On the slide-out panel, configure PVC parameters.

Click Create.

Parameter	Description
PVC Type	In this example, select EVS.
PVC Name	Enter the name of the PVC. After a PVC is created, a suffix is automatically added based on the number of pods. The format is <Custom PVC name>-<Serial number>, for example, example-0.
Creation Method	You can select Dynamically provision to create a PVC, PV, and underlying storage on the console in cascading mode.
Storage Classes	The default StorageClasses for EVS disks are csi-disk and csi-disk-topology. NOTE: If you use the csi-disk (EVS) StorageClass, a PVC and PV will be created immediately. The EVS disk is created with the PV, and then the PVC is bound to the PV. If you use the csi-disk-topology (EVS created with a delay) StorageClass, a PV will not be immediately created when a PVC is created. Instead, the pods that will be associated with the PVC are scheduled first, and then the EVS disk and PV are created, and finally the PV is bound to the PVC. You can customize a StorageClass and configure its reclaim policy and binding mode. For details, see Basic StorageClass Settings.
(Optional) Storage Volume Name Prefix	Available only when the cluster version is v1.23.14-r0, v1.25.9-r0, v1.27.6-r0, v1.28.4-r0, or later, and Everest v2.4.15 or later is installed in the cluster. This parameter specifies the name of the underlying storage that is automatically created. The actual underlying storage name is in the format of "Storage volume name prefix + PVC UID". If this parameter is left blank, the default prefix pvc will be used. For example, if the storage volume name prefix is set to test, the actual underlying storage name is test-{UID}.
AZ	Select the AZ of the EVS disk. The AZ must be the same as that of the cluster node. NOTE: An EVS disk can only be mounted to a node in the same AZ. After an EVS disk is created, its AZ cannot be changed.
Disk Type	Select an EVS disk type. EVS disk types vary depending on regions. Obtain the available EVS types on the console.
Capacity (GiB)	Capacity of the requested storage volume.
Access Mode	EVS volumes support only ReadWriteOnce, indicating that a storage volume can be mounted to one node in read/write mode. For details, see Volume Access Modes.
Encryption	Configure whether to encrypt the underlying storage. If you select Enable, select the target key or enter the key ID. Before using encryption, check whether the region where the EVS disk is located supports disk encryption.
Enterprise Project	An enterprise project manages multiple resources by category. Resources and projects across cloud service regions can be grouped into a single enterprise project. For example, an enterprise can classify resources by department or project group, and place relevant resources in the same enterprise project for unified management. This parameter is available only for enterprise accounts with enterprise projects enabled.
Resource Tag	You can add resource tags to classify resources. This function is supported only when the Everest version in the cluster is 2.1.39 or later. You can create predefined tags on the TMS console. These tags are available to all resources that support tags. You can use these tags to improve the tag creation and resource migration efficiency. CCE automatically creates system tags CCE-Cluster-ID=<cluster-ID>, CCE-Cluster-Name=<cluster-name>, and CCE-Namespace=<namespace-name>. These tags cannot be modified. NOTE: After a dynamic PV of the EVS type is created, the resource tags cannot be updated on the CCE console. To update these resource tags, go to the EVS console.

Enter the path to which the volume is mounted.

**Table 1** Mounting a storage volume
Parameter	Description
Mount Path	Enter a mount path, for example, /tmp. This parameter specifies a container path to which a data volume will be mounted. Do not mount the volume to a system directory such as / or /var/run. This may lead to container errors. Mount the volume to an empty directory. If the directory is not empty, ensure that there are no files that affect container startup. Otherwise, the files will be replaced, leading to container startup failures or workload creation failures. NOTICE: If a volume is mounted to a high-risk directory, use an account with minimum permissions to start the container. Otherwise, high-risk files on the host may be damaged.
Subpath	Enter the subpath of the storage volume and mount a path in the storage volume to the container. In this way, different folders of the same storage volume can be used in a single pod. tmp, for example, indicates that data in the mount path of the container is stored in the tmp folder of the storage volume. If this parameter is left blank, the root path will be used by default.
Permission	Read-only: You can only read the data in the mounted volumes. Read-write: You can modify the data volumes mounted to the path. Newly written data will not be migrated if the container is migrated, which may cause data loss.

In this example, the disk is mounted to the /data path of the container. The container data generated in this path is stored in the EVS disk.

Click to enlarge

Dynamically mount and use storage volumes. For details about other parameters, see Creating a StatefulSet. After the configuration, click Create Workload.

After the workload is created, the data in the container mount directory will be persistently stored. Verify the storage by referring to Verifying Data Persistence.

Use kubectl to access the cluster.

Create a file named statefulset-evs.yaml. In this example, the EVS volume is mounted to the /data path.

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: statefulset-evs
  namespace: default
spec:
  selector:
    matchLabels:
      app: statefulset-evs
  template:
    metadata:
      labels:
        app: statefulset-evs
    spec:
      containers:
        - name: container-1
          image: nginx:latest          # Replace it with the correct image path for your application.
          volumeMounts:
            - name: pvc-disk           # The value must be the same as that in the volumeClaimTemplates field.
              mountPath: /data         # Location where the storage volume is mounted
      imagePullSecrets:
        - name: default-secret
  serviceName: statefulset-evs         # Headless Service name
  replicas: 2
  volumeClaimTemplates:
    - apiVersion: v1
      kind: PersistentVolumeClaim
      metadata:
        name: pvc-disk
        namespace: default
        annotations:
          everest.io/disk-volume-type: SAS    # EVS disk type
          everest.io/crypt-key-id: <your_key_id>    # (Optional) Encryption key ID. Mandatory for an encrypted disk.

          everest.io/enterprise-project-id: <your_enterprise_project_id>  # (Optional) Enterprise project ID of the new EVS disk
          everest.io/disk-volume-tags: '{"key1":"value1","key2":"value2"}' # (Optional) Custom resource tags

          everest.io/csi.volume-name-prefix: test  # (Optional) Volume name prefix of the automatically-created underlying storage
        labels:
          failure-domain.beta.kubernetes.io/region: <your_region>   # Region of the node where the application is to be deployed
          failure-domain.beta.kubernetes.io/zone: <your_zone>       # AZ of the node where the application is to be deployed
      spec:
        accessModes:
          - ReadWriteOnce               # The value must be ReadWriteOnce for EVS disks.
        resources:
          requests:
            storage: 10Gi             # EVS disk capacity, ranging from 1 to 32768
        storageClassName: csi-disk    # The StorageClass is EVS.
---
apiVersion: v1
kind: Service
metadata:
  name: statefulset-evs   # Headless Service name
  namespace: default
  labels:
    app: statefulset-evs
spec:
  selector:
    app: statefulset-evs
  clusterIP: None
  ports:
    - name: statefulset-evs
      targetPort: 80
      nodePort: 0
      port: 80
      protocol: TCP
  type: ClusterIP

**Table 2** Key parameters
Parameter	Mandatory	Description
failure-domain.beta.kubernetes.io/region	Yes	Region where the cluster is located.
failure-domain.beta.kubernetes.io/zone	Yes	AZ where the EVS volume is created. It must be the same as the AZ planned for the workload.
everest.io/disk-volume-type	Yes	EVS disk type. All letters are in uppercase. SAS: High I/O SSD: Ultra-high I/O GPSSD: General Purpose SSD ESSD: Extreme SSD
everest.io/crypt-key-id	No	Mandatory when the EVS disk is encrypted. Enter the encryption key ID selected during EVS disk creation. To obtain an encryption key ID, log in to the Cloud Server Console. In the navigation pane, choose Elastic Volume Service > Disks. Click the name of the target EVS disk to go to its details page. On the Summary tab, copy the value of KMS Key ID in the Configuration Information area.
everest.io/enterprise-project-id	No	The ID of the enterprise project in which the EVS disk is created. It is applicable only to enterprise accounts with enterprise projects enabled. To obtain an enterprise project ID, log in to the EPS console, click the name of the target enterprise project, and copy the enterprise project ID.
everest.io/disk-volume-tags	No	(Optional) This parameter is supported when the Everest version in the cluster is 2.1.39 or later. You can add resource tags to classify resources. You can create predefined tags on the TMS console. These tags are available to all resources that support tags. You can use these tags to improve the tag creation and resource migration efficiency. CCE automatically creates system tags CCE-Cluster-ID=<cluster-ID>, CCE-Cluster-Name=<cluster-name>, and CCE-Namespace=<namespace-name>. These tags cannot be modified.
storage	Yes	Requested PVC capacity, in Gi. The value is an integer ranging from 1 to 32768. If storage is set to a decimal, the value will be rounded up for the EVS disk. For example, if storage is set to 10.1Gi, an 11-GiB EVS disk will be created.
storageClassName	Yes	StorageClass name, which is csi-disk for an EVS disk.

Run the following command to create a workload to which the EVS volume is mounted:
```
kubectl apply -f statefulset-evs.yaml
```
After the workload is created, the data in the container mount directory will be persistently stored. Verify the storage by referring to Verifying Data Persistence.

Verifying Data Persistence

View the deployed application and EVS volume files.

Run the following command to view the created pod:

kubectl get pod | grep statefulset-evs

Expected output:

statefulset-evs-0          1/1     Running   0             45s
statefulset-evs-1          1/1     Running   0             28s

Run the following command to check whether the EVS volume has been mounted to the /data path:

kubectl exec statefulset-evs-0 -- df | grep data

Expected output:

/dev/sdd              10255636     36888  10202364   0% /data

Run the following command to check the files in the /data path:
```
kubectl exec statefulset-evs-0 -- ls /data
```
Expected output:
```
lost+found
```

Run the following command to create a file named static in the /data path:
```
kubectl exec statefulset-evs-0 --  touch /data/static
```
Run the following command to check the files in the /data path:
```
kubectl exec statefulset-evs-0 -- ls /data
```
Expected output:
```
lost+found
static
```
Run the following command to delete the pod named statefulset-evs-0:
```
kubectl delete pod statefulset-evs-0
```
Expected output:
```
pod "statefulset-evs-0" deleted
```
After the deletion, the StatefulSet controller automatically creates a replica with the same name. Run the following command to check whether the files in the /data path have been modified:
```
kubectl exec statefulset-evs-0 -- ls /data
```
Expected output:
```
lost+found
static
```
The static file is retained, indicating that the data in the EVS volume can be stored persistently.

Related Operations

You can also perform the operations listed in Table 3.

**Table 3** Related operations
Operation	Description	Procedure
Expanding the capacity of an EVS disk	Quickly expand the capacity of a mounted EVS volume on the CCE console. Only the capacity of pay-per-use EVS disks can be expanded on the CCE console. To expand the capacity of yearly/monthly EVS disks, click the volume name to go to the EVS console.	Choose Storage in the navigation pane. In the right pane, click the PVCs tab. Locate the target PVC and choose More > Scale-out in the Operation column. Enter the capacity to be added and click OK.
Viewing events	View event names, event types, number of occurrences, Kubernetes events, first occurrence time, and last occurrence time of the PVC or PV.	Choose Storage in the navigation pane. In the right pane, click the PVCs or PVs tab. Locate the target PVC or PV and click View Events in the Operation column to view events generated within one hour (events are retained for one hour). You can also enable event persistence storage.
Viewing a YAML file	View, copy, or download the YAML file of a PVC or PV.	Choose Storage in the navigation pane. In the right pane, click the PVCs or PVs tab. Locate the target PVC or PV and click View YAML in the Operation column to view or download the YAML.

Helpful Links

If a storage volume fails to mount to a workload, rectify the fault by referring to What Should I Do If a Storage Volume Cannot Be Mounted or the Mounting Times Out?

Parent Topic: EVS

Previous topic: Using an EVS Disk Through a Dynamic PV

Next topic: Encrypting EVS Disks