Mount a static CPFS volume for persistent storage-Container Compute Service-阿里云帮助中心

Cloud Parallel File Storage (CPFS) is a fully managed parallel file system for high-performance computing (HPC) scenarios that supports on-demand extension. CPFS supports concurrent access from hundreds to thousands of machines, with tens of GB of throughput, millions of IOPS, and sub-millisecond latency. Mount a CPFS static volume to an ACS pod to provide your applications with shared, persistent, high-performance storage.

Usage notes

CPFS provides two editions: general-purpose CPFS and CPFS for Lingjun. Select an edition based on your business scenario and storage performance requirements. For the common scenarios and the differences between the two editions, see CPFS product overview.

Before you mount a volume, note the following items:

Invitational preview: CPFS for Lingjun supports end-to-end Remote Direct Memory Access (RDMA) networks. This edition is available only in specific regions and zones.
RDMA network requirement: When you access CPFS for Lingjun over an RDMA network, the pod and the file system must be in the same hpn-zone.
Shared storage: One CPFS volume can be mounted to multiple pods.
Supported compute classes: The following table lists the ACS pod compute classes that CPFS supports.

Specification	Access method	CPU type	GPU type
General-purpose CPFS	Virtual Private Cloud (VPC)	Supported	Supported
CPFS for Lingjun	VPC	Supported	Supported
CPFS for Lingjun	RDMA	Not supported	Supported by specific GPU models. For the GPU models that support the RDMA protocol, see Supported GPU types

Prerequisites

The managed-csiprovisioner add-on is installed in the ACS cluster.

Go to the ACS cluster management page in the ACS console. In the left-side navigation pane, click Components and Add-ons. On the Storage tab, verify that managed-csiprovisioner is installed.

Create a CPFS file system

General-purpose CPFS

Create a general-purpose CPFS file system.
After the file system is created, record the file system ID.
Create a Protocol Service.
Use the VPC and the vSwitch of the ACS cluster to create the Protocol Service and generate a directory-level mount target. After the Protocol Service is created, record the mount target domain name of the export directory. The domain name is in the cpfs-***-***.<Region>.cpfs.aliyuncs.com format.
On the file system details page, click Protocol Service in the left-side navigation pane. On the export directory list page, click the copy icon to the right of the mount address to copy the mount target domain name.

CPFS for Lingjun

Create a CPFS for Lingjun file system.
After the file system is created, record the file system ID.
(Optional) Create a VPC mount target.
For pods that do not support the RDMA protocol, such as CPU pods and some GPU pods, create a VPC mount target to access CPFS over the VPC.
Use the Virtual Private Cloud (VPC) and the vSwitch of the ACS cluster to create the VPC mount target and generate a mount address. After the mount target is created, record the mount target domain name. The domain name is in the cpfs-***-vpc-***.<Region>.cpfs.aliyuncs.com format.
On the file system details page, click Mounting in the left-side navigation pane. On the Mount Target tab, find the VPC mount target that you created. Click the view details icon to the right of the mount address to get the complete mount address and copy it.

Mount a CPFS volume

Step 1: Create a PV and a PVC

kubectl

Save the following YAML content to a file named cpfs-pv-pvc.yaml.

Select the YAML content based on the CPFS edition that you use. Also select the YAML based on the compute class of the pods to be mounted.

General-purpose CPFS

apiVersion: v1
kind: PersistentVolume
metadata:
  name: cpfs-test
  labels:
    alicloud-pvname: cpfs-test
spec:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Ti
  csi:
    driver: nasplugin.csi.alibabacloud.com
    volumeAttributes:
      mountProtocol: cpfs-nfs
      server: cpfs-***-***.cn-wulanchabu.cpfs.aliyuncs.com
      path: /share
    volumeHandle: cpfs-*****
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: cpfs-test
spec:
  accessModes:
  - ReadWriteMany
  selector:
    matchLabels:
      alicloud-pvname: cpfs-test
  resources:
    requests:
      storage: 10Ti

PV parameters:

Parameter	Description
`labels`	Sets a label so that the PVC can use `selector` to match and bind the PV.
`accessModes`	The access mode of the PV.
`capacity.storage`	The declared capacity of the volume.
`csi.driver`	The driver type. Set the value to `nasplugin.csi.alibabacloud.com`.
`csi.volumeAttributes`	The attributes of the CPFS volume. Set `mountProtocol` to `cpfs-nfs`. Set `server` to the mount target domain name of the export directory of the general-purpose CPFS Protocol Service. Set `path` to the path of the export directory of the general-purpose CPFS Protocol Service, such as `/share`. Optionally, specify a subdirectory, such as `/share/dir`.
`csi.volumeHandle`	The ID of the general-purpose CPFS file system.

PVC parameters:

Parameter	Description
`accessModes`	The access mode that the PVC requests from the PV.
`selector`	Matches and binds the PV based on the label on the PV.
`resources.requests.storage`	The storage capacity allocated to the pod. The value must not exceed the capacity of the PV.

CPFS for Lingjun

Important

CPFS for Lingjun can be mounted only to pods of specific GPU models. For the GPU models that support the RDMA protocol, see Supported GPU types.

apiVersion: v1
kind: PersistentVolume
metadata:
  name: cpfs-test
  labels:
    alicloud-pvname: cpfs-test
spec:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Ti
  csi:
    driver: bmcpfsplugin.csi.alibabacloud.com
    volumeAttributes:
      mountpointAutoSwitch: "true" # Enables automatic switchover between the RDMA mount address and the VPC mount address
      vpcMountTarget: cpfs-***-vpc-***.<Region>.cpfs.aliyuncs.com
      filesystemId: bmcpfs-*****
      path: /
    volumeHandle: bmcpfs-*****
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: cpfs-test
spec:
  accessModes:
  - ReadWriteMany
  selector:
    matchLabels:
      alicloud-pvname: cpfs-test
  resources:
    requests:
      storage: 10Ti

PV parameters:

Parameter	Description
`labels`	Sets a label so that the PVC can use `selector` to match and bind the PV.
`accessModes`	The access mode of the PV.
`capacity.storage`	The declared capacity of the volume.
`csi.driver`	The driver type. Set the value to `bmcpfsplugin.csi.alibabacloud.com`.
`csi.volumeAttributes`	The attributes of the CPFS volume. `filesystemId`: the ID of the CPFS for Lingjun file system. `path`: the default value `/` mounts the root directory of the CPFS file system. Optionally, specify a subdirectory, such as `/dir`. If the subdirectory does not exist, it is automatically created when the volume is mounted.
`csi.volumeHandle`	The ID of the CPFS for Lingjun file system.

PVC parameters:

Parameter	Description
`accessModes`	The access mode that the PVC requests from the PV.
`selector`	Matches and binds the PV based on the label on the PV.
`resources.requests.storage`	The storage capacity allocated to the pod. The value must not exceed the capacity of the PV.

Create the PV and the PVC.
```
kubectl create -f cpfs-pv-pvc.yaml
```

Verify that the PVC is bound to the PV.

kubectl get pvc cpfs-test

The following output is returned:

NAME        STATUS   VOLUME           CAPACITY   ACCESS MODES   STORAGECLASS    VOLUMEATTRIBUTESCLASS   AGE
cpfs-test   Bound    cpfs-test        10Ti       RWX            <unset>         <unset>                 10s

Console

Log on to the ACS console.
On the Clusters page, click the name of the cluster to go to the cluster management page.
In the left-side navigation pane of the cluster management page, choose Volumes > Persistent Volume Claims.
On the Persistent Volume Claims page, click Create.

In the dialog box that appears, configure the parameters and then click Create.

The following table describes the parameters for creating a PV together with a PVC. Alternatively, create the PV first and then create the PVC.

Note

In the console, mounting CPFS for Lingjun for CPU applications through a VPC mount target is currently not supported.

Parameter	Description	Example
PVC Type	Select CPFS.	CPFS
Name	The name of the PVC. Enter a custom name. For the format requirements, see the tips on the page.	cpfs-test
Allocation Mode	Select Existing Volumes or Create PV based on your requirements.	Create PV
CPFS Type	Select CPFS for LINGJUN or CPFS General-purpose based on your requirements. Select .	CPFS for LINGJUN
Access Mode	Both ReadWriteMany and ReadWriteOnce are supported.	ReadWriteMany
Mount Target Domain Name:	If you use general-purpose CPFS, specify the CPFS file system directory that you want to mount. If you enter only the mount address, such as `cpfs-*-.cn-wulanchabu.cpfs.aliyuncs.com`, the root directory (`/`) of the CPFS file system is mounted. If you enter the mount address and a subdirectory, such as `cpfs--*.cn-wulanchabu.cpfs.aliyuncs.com:/dir`, the `/dir` directory of the CPFS file system is mounted. If the `/dir` directory does not exist, it is automatically created.	`cpfs-*-*.cn-wulanchabu.cpfs.aliyuncs.com`
File System ID:	If you use CPFS for Lingjun, specify the ID of the CPFS for Lingjun file system that you want to mount.	`bmcpfs-0115******13q5`

View the PVC and the PV that you created.
On the Persistent Volume Claims page and the Persistent Volumes page, the new PVC and PV appear. Verify that they are bound.

Step 2: Create an application and mount CPFS

kubectl

Save the following YAML content to a file named cpfs-test.yaml.

GPU application

The following YAML example creates a Deployment that contains two pods. Both pods use the alibabacloud.com/compute-class: gpu label to declare that GPU computing power is required, request storage resources through the PVC named cpfs-test, and mount the volume to the /data path.

Note

For the specific GPU models, see Specify GPU models and driver versions for ACS GPU-accelerated pods.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: cpfs-test
  labels:
    app: cpfs-test
spec:
  replicas: 2
  selector:
    matchLabels:
      app: cpfs-test
  template:
    metadata:
      labels:
        app: cpfs-test
        # Specifies the GPU compute class
        alibabacloud.com/compute-class: gpu
        # Specifies the GPU model. Set this value based on your actual model, such as T4.
        alibabacloud.com/gpu-model-series: T4
        alibabacloud.com/compute-qos: default
    spec:
      containers:
      - name: nginx
        image: registry.cn-hangzhou.aliyuncs.com/acs-sample/nginx:latest
        ports:
        - containerPort: 80
        volumeMounts:
          - name: pvc-cpfs
            mountPath: /data
      volumes:
        - name: pvc-cpfs
          persistentVolumeClaim:
            claimName: cpfs-test

CPU application

The following YAML example creates a Deployment that contains two pods. The pods request storage resources through the PVC named cpfs-test and mount the volume to the /data path.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: cpfs-test
  labels:
    app: cpfs-test
spec:
  replicas: 2
  selector:
    matchLabels:
      app: cpfs-test
  template:
    metadata:
      labels:
        app: cpfs-test
    spec:
      containers:
      - name: nginx
        image: registry.cn-hangzhou.aliyuncs.com/acs-sample/nginx:latest
        ports:
        - containerPort: 80
        volumeMounts:
          - name: pvc-cpfs
            mountPath: /data
      volumes:
        - name: pvc-cpfs
          persistentVolumeClaim:
            claimName: cpfs-test

Create the Deployment and mount CPFS.
```
kubectl create -f cpfs-test.yaml
```

Check the status of the pods in the Deployment.

kubectl get pod | grep cpfs-test

The following output is returned. Two pods are created.

cpfs-test-****-***a   1/1     Running   0          45s
cpfs-test-****-***b   1/1     Running   0          45s

View the mount path.
The following command is an example. It lists the contents of the mount directory of the CPFS file system. For a new file system, the directory is empty by default, and no output is returned.
```
kubectl exec cpfs-test-****-***a -- ls /data
```

Console

In the left-side navigation pane of the cluster management page, choose Workloads > Deployments.
On the Deployments page, click Create from Image.

Configure the Deployment parameters and click Create.

The following tables describe the parameters that require your attention. Use the default values for the other parameters. For more information, see Create a stateless workload Deployment.

GPU application

Configuration page	Parameter	Description	Example
Basic Information	Application Name	The name of the Deployment. Enter a custom name. For the format requirements, see the tips on the page.	cpfs-test
	Replicas:	The number of replicas of the Deployment.	2
	Instance Type	Select the compute class of the pods. Note For the specific GPU models, see Specify GPU models and driver versions for ACS GPU-accelerated pods.	GPU, T4
Container	Image Name	Enter the address of the image used to deploy the application.	registry.cn-hangzhou.aliyuncs.com/acs-sample/nginx:latest
	Required Resources	Set the required GPU, vCPU, and memory resources.	GPU: 1. CPU: 2 vCPUs. Memory: 2 GiB.
	Volume	Click Add PVC and configure the parameters. Mount Source: select the PVC that you created. Container Path: enter the container path to which you want to mount the CPFS file system.	Mount Source: pvc-cpfs. Container Path: /data

CPU application

Configuration page	Parameter	Description	Example
Basic Information	Application Name	The name of the Deployment. Enter a custom name. For the format requirements, see the tips on the page.	cpfs-test
	Replicas:	The number of replicas of the Deployment.	2
	Instance Type	Select the compute class of the pods.	CPU, general-purpose
Container	Image Name	Enter the address of the image used to deploy the application.	registry.cn-hangzhou.aliyuncs.com/acs-sample/nginx:latest
	Required Resources	Set the required vCPU and memory resources.	CPU: 0.25 vCPUs. Memory: 0.5 GiB.
	Volume	Click Add PVC and configure the parameters. Mount Source: select the PVC that you created. Container Path: enter the container path to which you want to mount the CPFS file system.	Mount Source: pvc-cpfs. Container Path: /data

Check the status of the application.
1. On the Deployments page, click the application name.
2. On the Pods tab, verify that the pods are running. The status is Running.

Verify the shared storage and persistent storage of CPFS

The Deployment created in the preceding example contains two pods that mount the same CPFS file system. Verify the storage capabilities in the following ways:

Create a file in one pod and then view the file in the other pod to verify shared storage.
Restart the Deployment and then check whether the data in the file system still exists in the new pods to verify persistent storage.

View the pod information.

kubectl get pod | grep cpfs-test

The following output is returned:

cpfs-test-****-***a   1/1     Running   0          45s
cpfs-test-****-***b   1/1     Running   0          45s

Verify shared storage.
Create a file in one pod.
The following example uses the pod named cpfs-test-****-***a:
```
kubectl exec cpfs-test-****-***a -- touch /data/test.txt
```
View the file in the other pod.
The following example uses the pod named cpfs-test-****-***b:
```
kubectl exec cpfs-test-****-***b -- ls /data
```
The following output is returned. The new file test.txt is shared.
```
test.txt
```
Verify persistent storage.
Restart the Deployment to recreate its pods.
```
kubectl rollout restart deploy cpfs-test
```

View the pods and wait until the new pods are created.

kubectl get pod | grep cpfs-test

The following output is returned:

cpfs-test-****-***c   1/1     Running   0          78s
cpfs-test-****-***d   1/1     Running   0          52s

Check whether the data in the file system still exists in a new pod.
The following example uses the pod named cpfs-test-****-***c:
```
kubectl exec cpfs-test-****-***c -- ls /data
```
The following output is returned. The data in the CPFS file system still exists and is available in the mount directory of the new pod.
```
test.txt
```