Use CPFS volumes in ACK clusters - User Guide for Kubernetes Clusters| Alibaba Cloud Documentation Center

You can use Cloud Paralleled File System (CPFS) volumes in Container Service for Kubernetes (ACK) clusters. This topic describes how to install the FlexVolume plug-in in ACK clusters and use CPFS volumes in pods.

Prerequisites

A dedicated Kubernetes cluster is created. For more information, see Create a managed Kubernetes cluster.
The ACK cluster can be accessed from the Internet through Secure Shell (SSH). For more information, see Use SSH to connect to an ACK cluster.

Notice ACK supports only CPFS 1.0. CPFS 2.0 is not supported.

Background information

CPFS is a parallel file system. CPFS stores data across multiple data nodes in a cluster and allows data to be simultaneously accessed by multiple clients. Therefore, CPFS can provide data storage services with high input/output operations per second (IOPS), high throughput, and low latency for large-sized and high-performance computing clusters. CPFS is a shared storage service that meets the requirements on resource sharing and high performance. We recommend that you use ACK with CPFS in scenarios such as big data, artificial intelligence (AI), and genetic computing. For more information about CPFS, see What is CPFS?.

Step 1: Install drivers

To use CPFS volumes in an ACK cluster, you must install the following drivers:

CPFS container driver: the flexvolume-cpfs plug-in that is compatible with all CentOS versions. You can deploy flexvolume-cpfs to install the CPFS container driver.
CPFS client driver: the driver of the CPFS client. It is similar to nfs-client. This driver is heavily reliant on the operating system kernel. You can install the CPFS client driver by using the following methods:
- Manually install the driver. For more information, see Mount a file system.
- When you deploy flexvolume-cpfs, the CPFS client driver is automatically installed. However, the driver does not support all operating system kernels.
  You can run the uname -a command on a node to query the kernel version of the operating system. You can install the CPFS client driver on the node that uses one of the following kernel versions:
```
3.10.0-957.5.1
3.10.0-957.21.3
3.10.0-1062.9.1
```
Note
- FlexVolume allows you to only install the CPFS client driver (cpfs-client). After you install the driver on a node, you cannot reinstall and upgrade the driver.
- When you upgrade FlexVolume, only flexvolume-cpfs (CPFS container driver) is upgraded. cpfs-client (CPFS client driver) is not upgraded.
- When you install flexvolume-cpfs on nodes where cpfs-client and lustre are deployed, cpfs-client of the latest version is not automatically installed.
- You can only manually upgrade cpfs-client. For more information, see Mount a file system.

Deploy a YAML template on a node.

Use kubectl to connect to an ACK cluster from a client.
Create a flexvolume-cpfs.yaml file.

Copy the following content to the file:

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: flexvolume-cpfs
  namespace: kube-system
  labels:
    k8s-volume: flexvolume-cpfs
spec:
  selector:
    matchLabels:
      name: flexvolume-cpfs
  template:
    metadata:
      labels:
        name: flexvolume-cpfs
    spec:
      hostPID: true
      hostNetwork: true
      tolerations:
      - operator: "Exists"
      priorityClassName: system-node-critical
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
            - matchExpressions:
              - key: type
                operator: NotIn
                values:
                - virtual-kubelet
      containers:
      - name: flexvolume-cpfs
        image: registry.cn-hangzhou.aliyuncs.com/acs/flexvolume:v1.14.8.96-0d85fd1-aliyun
        imagePullPolicy: Always
        securityContext:
          privileged: true
        env:
        - name: ACS_CPFS
          value: "true"
        - name: FIX_ISSUES
          value: "false"
        livenessProbe:
          exec:
            command:
            - sh
            - -c
            - ls /acs/flexvolume
          failureThreshold: 8
          initialDelaySeconds: 15
          periodSeconds: 60
          successThreshold: 1
          timeoutSeconds: 15
        volumeMounts:
        - name: usrdir
          mountPath: /host/usr/
        - name: etcdir
          mountPath: /host/etc/
        - name: logdir
          mountPath: /var/log/alicloud/
        - mountPath: /var/lib/kubelet
          mountPropagation: Bidirectional
          name: kubeletdir
      volumes:
      - name: usrdir
        hostPath:
          path: /usr/
      - name: etcdir
        hostPath:
          path: /etc/
      - name: logdir
        hostPath:
          path: /var/log/alicloud/
      - hostPath:
          path: /var/lib/kubelet
          type: Directory
        name: kubeletdir
  updateStrategy:
    rollingUpdate:
      maxUnavailable: 10%
    type: RollingUpdate

Run the following command to deploy the YAML file on the node:
```
kubectl create -f flexvolume-cpfs.yaml
```

Check the deployment result.

Run the following command to verify that the volume plug-ins are deployed:

kubectl get pod -nkube-system | grep flex

The following output is returned:

flexvolume-97psk                                  1/1     Running   0          27m
flexvolume-cpfs-dgxfq                             1/1     Running   0          98s
flexvolume-cpfs-qpbcb                             1/1     Running   0          98s
flexvolume-cpfs-vlrf9                             1/1     Running   0          98s
flexvolume-cpfs-wklls                             1/1     Running   0          98s
flexvolume-cpfs-xtl9b                             1/1     Running   0          98s
flexvolume-j8zjr                                  1/1     Running   0          27m
flexvolume-pcg4l                                  1/1     Running   0          27m
flexvolume-tjxxn                                  1/1     Running   0          27m
flexvolume-x7ljw                                  1/1     Running   0          27m

Note The flexvolume-cpfs plug-in is deployed on pods whose names are prefixed with flexvolume-cpfs. Pods whose names do not contain cpfs are deployed with the FlexVolume plug-in. Alibaba Cloud disks, Apsara File Storage NAS (NAS) file systems, and Object Storage Service (OSS) buckets can be mounted to these pods. Both types of plug-ins can be deployed on the same node.

Run the following command to check whether cpfs-client is installed:
```
rpm -qa | grep cpfs
```
The following output is returned:
```
kmod-cpfs-client-2.10.8-202.el7.x86_64
cpfs-client-2.10.8-202.el7.x86_64
```
Run the following command to check whether mount.lustre is installed:
```
which mount.lustre
```
The following output is returned:
```
/usr/sbin/mount.lustre
```

Step 2: Use a CPFS volume

To use a CPFS volume in an ACK cluster, you must first create a CPFS file system and a mount target in the CPFS console. For more information, see Create a file system.

Notice When you create a CPFS mount target, select a virtual private cloud (VPC). The ACK cluster and the mount target must be deployed in the same VPC.

In the following example, the mount target is cpfs-*-alup.cn-shenzhen.cpfs.nas.aliyuncs.com@tcp:cpfs--ws5v.cn-shenzhen.cpfs.nas.aliyuncs.com@tcp and the file system ID is 0237ef41.

Create a persistent volume (PV).

Create a pv-cpfs.yaml file.

Copy the following content to the file:

apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv-cpfs
  labels:
    alicloud-pvname: pv-cpfs
spec:
  capacity:
    storage: 5Gi
  accessModes:
    - ReadWriteMany
  flexVolume:
    driver: "alicloud/cpfs"
    options:
      server: "cpfs-****-alup.cn-shenzhen.cpfs.nas.aliyuncs.com@tcp:cpfs-***-ws5v.cn-shenzhen.cpfs.nas.aliyuncs.com@tcp"
      fileSystem: "0237ef41"
      subPath: "/k8s"
      options: "ro"


Parameter	Description
server	Set the value to the mount target of the CPFS file system.
fileSystem	Set the value to the ID of the CPFS file system.
subPath	Set the value to a subdirectory of the CPFS root.
options	Other mount options. This parameter is optional.

Run the following command to create a PV:
```
kubectl create -f pv-cpfs.yaml
```

Create a persistent volume claim (PVC).

Create a pvc-cpfs file and copy the following content to the file:

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: pvc-cpfs
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 5Gi
  selector:
    matchLabels:
      alicloud-pvname: pv-cpfs

Run the kubectl create -f pvc-cpfs command to create a PVC.

Create a Deployment.

Create a nas-cpfs file and copy the following content to the file:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nas-cpfs
  labels:
    app: nginx
spec:
  replicas: 1
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        ports:
        - containerPort: 80
        volumeMounts:
          - name: pvc-cpfs
            mountPath: "/data"
      volumes:
        - name: pvc-cpfs
          persistentVolumeClaim:
            claimName: pvc-cpfs

Run the kubectl create -f nas-cpfs command to create a Deployment.

Result

Run the following command to query the mounting states of the pods that run on the node:

kubectl get pod

The following output is returned:

NAME                        READY   STATUS    RESTARTS   AGE
nas-cpfs-79964997f5-kzrtp   1/1     Running   0          45s

Run the following command to query the directories that are mounted to pods on the node:

kubectl exec -ti nas-cpfs-79964997f5-kzrtp sh
mount | grep k8s

The following output is returned:

192.168.1.12@tcp:192.168.1.10@tcp:/0237ef41/k8s on /data type lustre (ro,lazystatfs)

Run the following command to query the mounted directories on the node:

mount | grep cpfs

The following output is returned:

192.168.1.12@tcp:192.168.1.10@tcp:/0237ef41/k8s on /var/lib/kubelet/pods/c4684de2-26ce-11ea-abbd-00163e12e203/volumes/alicloud~cpfs/pv-cpfs type lustre (ro,lazystatfs)