This is the multi-page printable view of this section. Click here to print.

Return to the regular view of this page.

Managing PVs

1: Expanding the Capacity of a PV
2: Cloning a PV
3: Changing a PV

3.1: Enabling the PVC Change Feature

3.1.1: Enabling the PVC Change Feature Using Helm
3.1.2: Enabling the PVC Change Feature Manually

3.2: Configuring PVC Changes

3.2.1: Creating a PVC Change
3.2.2: Querying a PVC Change
3.2.3: Deleting a PVC Change

1 - Expanding the Capacity of a PV

For OceanStor V700R001C10 as well as OceanStor Dorado V700R001C10 and later versions, the minimum capacity of a file system after capacity expansion is limited. For details, see the product documentation of the corresponding storage device.

When the capacity of a PVC used by a container is insufficient, you need to expand the capacity of the PVC.

Prerequisites

A PVC has been created, the backend to which it resides exists and supports capacity expansion.
Compatibility and Features lists the storage that supports capacity expansion, and Kubernetes Feature Matrix lists the Kubernetes versions that support capacity expansion.

The csi-resizer service is enabled for huawei-csi-controller.

kubectl describe deploy huawei-csi-controller -n huawei-csi | grep csi-resizer

If the following information is displayed, the csi-resizer service is enabled.

   csi-resizer:
    Image:      k8s.gcr.io/sig-storage/csi-resizer:v1.9.0

Procedure

Run the following command to check whether the StorageClass supports capacity expansion. In the preceding command, mysc indicates the name of the StorageClass to be queried.
```
kubectl get sc mysc
```
The following is an example of the command output.
```
NAME              PROVISIONER      RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
mysc              csi.huawei.com   Delete          Immediate           true                  172m
```
If the value of ALLOWVOLUMEEXPANSION is true, the current StorageClass supports capacity expansion. In this case, go to 3 .
Run the following command to change the value of allowVolumeExpansion to true. In the preceding command, mysc indicates the name of the StorageClass to be modified.
```
kubectl patch sc mysc --patch '{"allowVolumeExpansion":true}'
```

Run the following command to query the StorageClass name of the PVC. In the preceding command, mypvc indicates the name of the PVC to be expanded.

kubectl get pvc mypvc

The following is an example of the command output.

NAME               STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS      AGE
mypvc              Bound    pvc-3383be36-537c-4cb1-8f32-a415fa6ba384   2Gi        RW0            mysc              145m

Run the following command to expand the capacity.
```
kubectl patch pvc mypvc -p '{"spec":{"resources":{"requests":{"storage":"120Gi"}}}}'
```
In the preceding command, mypvc indicates the name of the PVC to be expanded, and 120Gi indicates the capacity after expansion. Change the values based on the site requirements.
- The PVC capacity depends on storage specifications and host specifications. For example, OceanStor Dorado 6.1.2 or OceanStor Pacific series 8.1.0 is connected to CentOS 7. If ext4 file systems are used, see Table 2 . If XFS file systems are used, see Table 3 . If NFS or raw devices are used, the capacity must meet the specifications of the used Huawei storage device model and version.
- If the PVC capacity does not meet the specifications, a PVC or Pod may fail to be created due to the limitations of storage specifications or host file system specifications.
- If the capacity expansion fails because the target capacity exceeds the storage pool capacity, see Failed to Expand the PVC Capacity Because the Target Capacity Exceeds the Storage Pool Capacity .

Run the following command to check whether the capacity modification takes effect.

kubectl get pvc

The following is an example of the command output. If the value of CAPACITY is changed to the specified capacity, the capacity expansion is successful.

NAME        STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS   AGE
mypvc       Bound    pvc-3383be36-537c-4cb1-8f32-a415fa6ba384   120Gi       RWO            mysc           24s

2 - Cloning a PV

This section describes how to clone a PVC.

When cloning a PVC, you need to specify the data source. The following is an example of cloning a PVC. In this example, mypvc is used as the data source and a PVC named myclone is created.

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: myclone
spec:
  storageClassName: mysc
  dataSource:
    name: mypvc
    kind: PersistentVolumeClaim
  volumeMode: Filesystem
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 2Gi

The specified storageClassName must be the same as the StorageClass of the source volume in dataSource.
The capacity of the clone volume must be greater than or equal to that of the source volume. Equal capacity is recommended.

Prerequisites

The source PVC already exists in the system, and the backend where the source PVC resides supports cloning. Compatibility and Features lists the storage that supports cloning, and Kubernetes Feature Matrix lists the Kubernetes versions that support cloning.

Procedure

Run the following command to create a PVC based on the configuration file of the clone volume.
```
kubectl create -f myclone.yaml
```

3 - Changing a PV

3.1 - Enabling the PVC Change Feature

The PVC change feature is disabled by default during Huawei CSI installation. To use this feature, perform the following steps.

3.1.1 - Enabling the PVC Change Feature Using Helm

Prerequisites

You have installed Huawei CSI using Helm.

Procedure

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.
Run the following command to check whether the PVC change feature is enabled.
helm-huawei-csi indicates the Helm chart name specified during installation, and huawei-csi indicates the Helm chart namespace specified during installation. For details about the component package path, see Table 1 .
```
helm get values helm-huawei-csi -n huawei-csi -a | grep volumeModify -A 1
```
The following is an example of the command output.
- If enabled: true is displayed in the command output, the feature is enabled. In this case, skip the following steps.
- If enabled: false is displayed in the command output, perform the following steps to enable the PVC change feature.
```
volumeModify:
  enabled: false
```
Go to the /helm/esdk directory and run the following command to configure the volume change CRD.
```
# kubectl apply -f ./crds/volume-modify/
customresourcedefinition.apiextensions.k8s.io/volumemodifyclaims.xuanwu.huawei.io configured
customresourcedefinition.apiextensions.k8s.io/volumemodifycontents.xuanwu.huawei.io configured
```
If the command output contains Warning: resource customresourcedefinitions/volumemodifycontents.xuanwu.huawei.io is missing the kubectl.kubernetes.io/last-applied-configuration…, you can ignore it. This message is displayed because the kubectl create command instead of the kubectl apply command is used for installation by Helm.
Run the following command to obtain the original service configuration file.
```
helm get values helm-huawei-csi -n huawei-csi -a > ./update-values.yaml
```
Run the vi update-values.yaml command to open the file obtained in 4 and modify the following configuration. After the modification is complete, press Esc and enter :wq! to save the modification.
```
csiExtender:
  volumeModify:    
    enabled: true
```

Run the following command to update Huawei CSI services.

helm upgrade helm-huawei-csi ./ -n huawei-csi  -f ./update-values.yaml

Run the following command to check whether the services are started.

kubectl get pod -n huawei-csi

The following is an example of the command output. In the preceding command, huawei-csi indicates the namespace for deploying Huawei CSI.

NAME                                     READY     STATUS    RESTARTS   AGE
huawei-csi-controller-6dfcc4b79f-9vjtq   10/10     Running   0          24m
huawei-csi-node-tqs87                    3/3       Running   0          20m

3.1.2 - Enabling the PVC Change Feature Manually

Prerequisites

Huawei CSI has been manually installed.

Procedure

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.
Go to the manual/esdk working directory and run the following command to configure the volume change CRD.
```
kubectl apply -f ./crds/volume-modify/
```
Run the following command. For details about the component package path, see Table 1 .
```
kubectl apply -f ./deploy/huawei-csi-controller-extender.yaml
```

Run the following command to check whether the services are started.

kubectl get pod -n huawei-csi

The following is an example of the command output. In the preceding command, huawei-csi indicates the namespace for deploying Huawei CSI.

NAME                                     READY     STATUS    RESTARTS   AGE
huawei-csi-controller-6dfcc4b79f-9vjtq   10/10     Running   0          24m
huawei-csi-node-tqs87                    3/3       Running   0          24m

3.2 - Configuring PVC Changes

The PVC change feature is implemented using CRD. Related resources are described as follows.

Table 1 Resource description

NAME	APIVERSION	NAMESPACED	KIND
volumemodifyclaims	xuanwu.huawei.io/v1	false	VolumeModifyClaim
volumemodifycontents	xuanwu.huawei.io/v1	false	VolumeModifyContent

VolumeModifyClaim resources can be created, deleted, and queried, but cannot be updated.
VolumeModifyContent resources can only be queried and are used to display the change details of a single PVC. Do not manually create, delete, or modify the resources.
VolumeModifyContent resources are managed by VolumeModifyClaim. Do not manually manage VolumeModifyContent resources.

3.2.1 - Creating a PVC Change

Prerequisites

The storage backends associated with the PVC to be changed are HyperMetro storage backends. If they are not HyperMetro storage backends, configure them by following the instructions in Updating a Storage Backend .

PVC Change File Description

The sample template of the PVC change file is /examples/volumemodifyclaim.yaml. The following table lists the configuration items.

Table 1 Parameter description

Parameter	Description	Mandatory	Default Value	Remarks
apiVersion	API group, which is of the string type.	Yes	xuanwu.huawei.io/v1	The value is fixed at xuanwu.huawei.io/v1.
kind	Resource type, which is of the string type.	Yes	VolumeModifyClaim	The value is fixed at VolumeModifyClaim.
metadata.name	Name of a cluster resource object, which is of the string type.	Yes	-	The name must comply with the naming rules of a DNS subdomain name. The value can contain a maximum of 63 characters, including digits, lowercase letters, hyphens (-), and periods (.). It must start and end with a lowercase letter or digit. Note: During a PVC change, the original StorageClass is backed up. The name of the backup StorageClass is <Original StorageClass name><VolumeModifyClaim name>, and must comply with the StorageClass naming rules.
spec.source.kind	Data source type, which is of the string type.	Yes	StorageClass	This parameter can only be set to StorageClass.
spec.source.name	Data source name, which is of the string type.	Yes	-	Only a StorageClass name can be configured.
spec.parameters.hyperMetro	Whether to change a common volume to a HyperMetro volume. Currently, the value can only be "true".	Yes	-	Only common storage volumes at the primary site can be changed to HyperMetro storage volumes.
spec.parameters.metroPairSyncSpeed	Data synchronization speed of a HyperMetro pair. The value ranges from 1 to 4. The value can be: 1: low 2: medium 3: high 4: highest	No	-	This parameter is available only when spec.parameters.hyperMetro is set to "true". Note: If this parameter is not configured, the storage speed of the HyperMetro pair is determined by the storage device. The highest synchronization speed may increase the host latency.

The spec.source.kind and spec.source.name parameters are used to specify the volume change scope. For example, if they are set to a StorageClass and the corresponding name respectively, all PVCs in the Bound state provisioned using the target StorageClass will be changed.
After all associated PVCs are changed, Huawei CSI will replace the original StorageClass and add the spec.parameters parameter of the VolumeModifyClaim so that the PVCs meet the StorageClass definition.

For details about the configuration in typical scenarios, see the following example:

Changing a Common Volume to a HyperMetro Volume

The following is an example of changing a common volume to a HyperMetro volume:

apiVersion: xuanwu.huawei.io/v1
kind: VolumeModifyClaim
metadata:
  name: myvmc
spec:
  source:
    kind: StorageClass
    name: mysc
  parameters:
    hyperMetro: "true"

Creating a PVC Change Resource

The changed HyperMetro volumes must be in HyperMetro AA mode.
When a common volume is changed to a HyperMetro volume, only the storage volume at the primary site can be changed.
Do not use Huawei CSI to manage a PVC during PVC change resource creation.
Multiple VolumeModifyClaim resources cannot be created for the same PVC. If the target PVC needs to be changed for multiple times, perform the changes one by one.

To create a PVC change resource using a PVC change file, perform the following steps:

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.
Run the following command to create a PVC change.
```
kubectl create -f volumemodifyclaim.yaml 
```
Query the creation result by following the instructions in Querying a PVC Change .

3.2.2 - Querying a PVC Change

This section describes how to use Kubectl to query the PVC change status. Currently, Huawei CSI provides the following APIs through CRD.

Querying a VolumeModifyClaim

To query a VolumeModifyClaim using kubectl, perform the following steps.

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.

Run the following command to query a PVC change. In the command, vmc-name indicates the name of the VolumeModifyClaim resource.

kubectl get volumemodifyclaims <vmc-name> -owide

The following is an example of the command output.

NAME    STATUS      READY   SOURCEKIND     SOURCENAME   STARTEDAT              COMPLETEDAT            AGE
myvmc   Completed   1/1     StorageClass   mysc         2024-06-06T03:19:13Z   2024-06-06T03:19:16Z   2m2s

Table 1 Command output description

Parameter	Description
NAME	VolumeModifyClaim resource name.
STATUS	VolumeModifyClaim resource status. The value can be: Pending: initial status. Creating: The VolumeModifyClaim has completed basic verification and the server has received the change task, but the task has not been completed. Completed: All associated PVCs are changed. Rollback: When associated PVCs are partially changed, a user deletes PVCs. Deleting: When all associated PVCs are changed, a user deletes PVCs.
READY	Ratio of the number of changed PVCs to the total number of PVCs that need to be changed.
SOURCEKIND	Data source type, for example, StorageClass.
SOURCENAME	Data source name, for example, StorageClass name.
STARTEDAT	Change start time, that is, the timestamp when the server receives the task and starts to process the task.
COMPLETEDAT	Change completion time, that is, the timestamp when the changes of all associated PVCs are complete. This parameter exists only when STATUS is Completed.
AGE	Lifetime of a VolumeModifyClaim from the time when it is created to the current time.

You can use kubectl to view the Events information of a VolumeModifyClaim. If a VolumeModifyClaim cannot meet the creation requirements or an error occurs during the creation, the server will record the Events information. The following command is used as an example:
kubectl describe volumemodifyclaims local-to-hypermetro 

Querying a VolumeModifyContent

A VolumeModifyContent is created using a VolumeModifyClaim and records the change details of a single PVC. To query a VolumeModifyContent using kubectl, perform the following steps:

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.

Run the following command to query a PVC change. In the command, myvmc-uid indicates the VolumeModifyContent resource name.

kubectl get volumemodifycontents myvmc-uid  -owide

The following is an example of the command output.

NAME         STATUS      MODIFYCLAIMNAME     SOURCEVOLUME   STARTEDAT              COMPLETEDAT            AGE
myvmc-uid    Completed   myvmc               default/mypvc  2024-06-06T03:19:07Z   2024-06-06T03:19:09Z   36m

Table 2 Command output description

Parameter	Description
NAME	VolumeModifyContent resource name. The format is VolumeModifyClaim name-UID of the associated PVC.
STATUS	VolumeModifyContent resource status. The value can be: Pending: initial status. Creating: The VolumeModifyContent has completed basic verification and the server has received the change task, but the task has not been completed. Completed: The associated PVC is changed. Rollback: The PVC change is being rolled back.
MODIFYCLAIMNAME	Name of the associated VolumeModifyClaim.
SOURCEVOLUME	Information about the associated PVC. The format is Namespace name/PVC name.
STARTEDAT	PVC change start time, that is, the timestamp when the server receives the task and starts to process the task.
COMPLETEDAT	PVC change completion time, that is, the timestamp when the changes of all associated PVCs are complete. This parameter exists only when STATUS is Completed.
AGE	Lifetime of a VolumeModifyContent from the time when it is created to the current time.

You can use kubectl to view the Events information of a VolumeModifyContent. If a VolumeModifyContent cannot meet the creation requirements or an error occurs during the PVC change, the server will record the Events information. The following command is used as an example:
kubectl describe volumemodifycontents myvmc-uid

3.2.3 - Deleting a PVC Change

If STATUS of a VolumeModifyClaim is Creating, deleting the VolumeModifyClaim resource will delete the created resource on the storage side and then remove the cluster resource. After the deletion, if you continue to use the original StorageClass for PVC management, you need to restore the associated storage backend to a non-HyperMetro storage backend.
If STATUS of a VolumeModifyClaim is Pending or Completed, deleting the VolumeModifyClaim resource will only remove the cluster resource and will not delete the created resource on the storage side (that is, there is not interaction with the storage side).
VolumeModifyContent resources are managed by VolumeModifyClaim. Do not manually manage VolumeModifyContent resources.
If some PVCs among the PVCs to be changed meet the change requirements and the batch change fails, all PVC changes will be removed. As a result, the PVCs that meet the change requirements will not meet the change requirements.
If a PVC to be changed has been manually managed on the storage side, the change may fail. Do not manually manage storage volumes when using the change feature.

This section describes how to use kubectl to delete a PVC change. The procedure is as follows.

Procedure

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.
Run the following command to delete a PVC change. In the command, vmc-name indicates the name of the VolumeModifyClaim resource.
```
kubectl delete volumemodifyclaims <vmc-name>
```
Query the creation result by following the instructions in Querying a PVC Change .