Mount a NAS file system to a sandboxed container through the CSI plug-in

How the direct mount solution works

Example

The following example describes how to mount a NAS file system to a sandboxed container. In this example, an Apsara File Storage NAS instance is created and a YAML file template is used to create resource objects.

1. Create an Apsara File Storage NAS instance. For more information, see Step 1: Create a file system.
   Notice The NAS instance must be deployed in the same VPC as the cluster.
   Obtain the mount point URL of the NAS instance. In the following figure, file-system-id.region.nas.aliyuncs.com in the Mount Command column is the URL of the mount point.

   Note file-system-id.region.nas.aliyuncs.com is the URL of the mount point. To obtain the mount point URL, log on to the NAS console, click the name of the target file system, and choose Mounting Usage.
2. Use the following template to create resource objects:
   Notice Replace ${nas-server-address} in the template with the URL of the mount point.

   cat <<EOF | kubectl create -f -
   apiVersion: v1
   kind: PersistentVolumeClaim
   metadata:
     name: nas-pvc-csi
     namespace: default
   spec:
     accessModes:
       - ReadWriteMany
     resources:
       requests:
         storage: 5Gi
     selector:
       matchLabels:
         alicloud-pvname: nas-pv-csi
   ---
   apiVersion: v1
   kind: PersistentVolume
   metadata:
     labels:
       alicloud-pvname: nas-pv-csi
     name: nas-pv-csi
   spec:
     accessModes:
       - ReadWriteMany
     capacity:
       storage: 5Gi
     csi:
       driver: nasplugin.csi.alibabacloud.com
       volumeAttributes:
         options: noresvport,nolock
         path: /csi
         server: ${nas-server-address}
         vers: "3"
       volumeHandle: nas-pv-csi
     persistentVolumeReclaimPolicy: Retain
   ---
   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: deploy-nas-csi
   spec:
     replicas: 2
     selector:
       matchLabels:
         app: busybox
     template:
       metadata:
         labels:
           app: busybox
       spec:
         runtimeClassName: runv
         containers:
           - name: busybox
             image: registry.cn-hangzhou.aliyuncs.com/acs/busybox:v1.29.2
             command: 
             - tail
             - -f
             - /dev/null
             volumeMounts:
               - name: nas-pvc
                 mountPath: "/data"
         restartPolicy: Always
         volumes:
           - name: nas-pvc
             persistentVolumeClaim:
               claimName: nas-pvc-csi
   EOF

3. Run the following commands to check the mount point type in the pod.

   kubectl get pods
   kubectl exec -it ${podid} sh
   mount | grep /data | grep nfs

   If the command output is not empty, the NAS file system is directly mounted to the container.

I/O performance comparison

• Random read operations per second

  Mount mode	Result
  Host	read: IOPS=24.3k, BW=94.8MiB/s (99.4MB/s)(5687MiB/60004msec)
  CSI plug-in	read: IOPS=26.0k, BW=102MiB/s (107MB/s)(6101MiB/60005msec)
  9PFS	read: IOPS=780, BW=3120KiB/s (3195kB/s)(183MiB/60001msec)

• Random write operations per second

  Mount mode	Result
  Host	write: IOPS=14.3k, BW=55.9MiB/s (58.6MB/s)(3354MiB/60006msec)
  CSI plug-in	write: IOPS=13.8k, BW=53.8MiB/s (56.4MB/s)(3226MiB/60008msec)
  9PFS	write: IOPS=356, BW=1428KiB/s (1462kB/s)(83.7MiB/60001msec)

• Random read throughput

  Mount mode	Result
  Host	read: IOPS=331, BW=331MiB/s (347MB/s)(29.9GiB/92439msec)
  CSI plug-in	read: IOPS=251, BW=252MiB/s (264MB/s)(29.9GiB/121561msec)
  9PFS	read: IOPS=130, BW=130MiB/s (136MB/s)(7861MiB/60422msec)

• Random write throughput

  Mount mode	Result
  Host	write: IOPS=507, BW=507MiB/s (532MB/s)(29.9GiB/60372msec)
  CSI plug-in	write: IOPS=508, BW=508MiB/s (533MB/s)(29.9GiB/60255msec)
  9PFS	write: IOPS=1, BW=1500KiB/s (1536kB/s)(88.0MiB/60061msec)

According to the preceding data, the throughput and IOPS are almost the same when the NAS file system is mounted to the container and when the NAS file system is mounted to the host. The I/O performance is worst when the NAS file system is mounted through 9PFS.