Use Azure Files Container Storage Interface (CSI) driver in Azure Kubernetes Service (AKS)

The Azure Files Container Storage Interface (CSI) driver is a CSI specification-compliant driver used by Azure Kubernetes Service (AKS) to manage the lifecycle of Azure file shares. The CSI is a standard for exposing arbitrary block and file storage systems to containerized workloads on Kubernetes.

By adopting and using CSI, AKS now can write, deploy, and iterate plug-ins to expose new or improve existing storage systems in Kubernetes. Using CSI drivers in AKS avoids having to touch the core Kubernetes code and wait for its release cycles.

Azure Files CSI driver new features

In addition to the original in-tree driver features, Azure Files CSI driver supports the following new features:

• Network File System (NFS) version 4.1
• Private endpoint
• Creating large mount of file shares in parallel.

Use a persistent volume with Azure Files

A persistent volume (PV) represents a piece of storage that's provisioned for use with Kubernetes pods. A PV can be used by one or many pods and can be dynamically or statically provisioned. If multiple pods need concurrent access to the same storage volume, you can use Azure Files to connect by using the Server Message Block (SMB) or NFS protocol. This article shows you how to dynamically create an Azure Files share for use by multiple pods in an AKS cluster. For static provisioning, see Manually create and use a volume with an Azure Files share.

With Azure Files shares, there is no limit as to how many can be mounted on a node.

For more information on Kubernetes volumes, see Storage options for applications in AKS.

Dynamically create Azure Files PVs by using the built-in storage classes

A storage class is used to define how an Azure file share is created. A storage account is automatically created in the node resource group for use with the storage class to hold the Azure files share. Choose one of the following Azure storage redundancy SKUs for skuName:

• Standard_LRS: Standard locally redundant storage
• Standard_GRS: Standard geo-redundant storage
• Standard_ZRS: Standard zone-redundant storage
• Standard_RAGRS: Standard read-access geo-redundant storage
• Standard_RAGZRS: Standard read-access geo-zone-redundant storage
• Premium_LRS: Premium locally redundant storage

When you use storage CSI drivers on AKS, there are two more built-in StorageClasses that uses the Azure Files CSI storage drivers. The other CSI storage classes are created with the cluster alongside the in-tree default storage classes.

• azurefile-csi: Uses Azure Standard Storage to create an Azure file share.
• azurefile-csi-premium: Uses Azure Premium Storage to create an Azure file share.

The reclaim policy on both storage classes ensures that the underlying Azure files share is deleted when the respective PV is deleted. The storage classes also configure the file shares to be expandable, you just need to edit the persistent volume claim (PVC) with the new size.

To use these storage classes, create a PVC and respective pod that references and uses them. A PVC is used to automatically provision storage based on a storage class. A PVC can use one of the pre-created storage classes or a user-defined storage class to create an Azure files share for the desired SKU and size. When you create a pod definition, the PVC is specified to request the desired storage.

Create an example PVC and pod that prints the current date into an outfile by running the kubectl apply commands:

kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/azurefile-csi-driver/master/deploy/example/pvc-azurefile-csi.yaml
kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/azurefile-csi-driver/master/deploy/example/nginx-pod-azurefile.yaml

The output of the command resembles the following example:

persistentvolumeclaim/pvc-azurefile created
pod/nginx-azurefile created

After the pod is in the running state, you can validate that the file share is correctly mounted by running the following command and verifying the output contains the outfile:

kubectl exec nginx-azurefile -- ls -l /mnt/azurefile

The output of the command resembles the following example:

total 29
-rwxrwxrwx 1 root root 29348 Aug 31 21:59 outfile

Create a custom storage class

The default storage classes suit the most common scenarios, but not all. For some cases, you might want to have your own storage class customized with your own parameters. For example, use the following manifest to configure the mountOptions of the file share.

The default value for fileMode and dirMode is 0777 for Kubernetes mounted file shares. You can specify the different mount options on the storage class object.

Create a file named azure-file-sc.yaml, and paste the following example manifest:

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: my-azurefile
provisioner: file.csi.azure.com
reclaimPolicy: Delete
volumeBindingMode: Immediate
allowVolumeExpansion: true
mountOptions:
  - dir_mode=0640
  - file_mode=0640
  - uid=0
  - gid=0
  - mfsymlinks
  - cache=strict # https://linux.die.net/man/8/mount.cifs
  - nosharesock
parameters:
  skuName: Standard_LRS

Create the storage class by running the kubectl apply command:

kubectl apply -f azure-file-sc.yaml

The output of the command resembles the following example:

storageclass.storage.k8s.io/my-azurefile created

The Azure Files CSI driver supports creating snapshots of persistent volumes and the underlying file shares.

Create a volume snapshot class with the kubectl apply command:

kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/azurefile-csi-driver/master/deploy/example/snapshot/volumesnapshotclass-azurefile.yaml

The output of the command resembles the following example:

volumesnapshotclass.snapshot.storage.k8s.io/csi-azurefile-vsc created

Create a volume snapshot from the PVC we dynamically created at the beginning of this tutorial, pvc-azurefile.

kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/azurefile-csi-driver/master/deploy/example/snapshot/volumesnapshot-azurefile.yaml

The output of the command resembles the following example:

volumesnapshot.snapshot.storage.k8s.io/azurefile-volume-snapshot created

Verify the snapshot was created correctly by running the following command:

kubectl describe volumesnapshot azurefile-volume-snapshot

The output of the command resembles the following example:

Name:         azurefile-volume-snapshot
Namespace:    default
Labels:       <none>
Annotations:  API Version:  snapshot.storage.k8s.io/v1beta1
Kind:         VolumeSnapshot
Metadata:
  Creation Timestamp:  2020-08-27T22:37:41Z
  Finalizers:
    snapshot.storage.kubernetes.io/volumesnapshot-as-source-protection
    snapshot.storage.kubernetes.io/volumesnapshot-bound-protection
  Generation:        1
  Resource Version:  955091
  Self Link:         /apis/snapshot.storage.k8s.io/v1beta1/namespaces/default/volumesnapshots/azurefile-volume-snapshot
  UID:               c359a38f-35c1-4fb1-9da9-2c06d35ca0f4
Spec:
  Source:
    Persistent Volume Claim Name:  pvc-azurefile
  Volume Snapshot Class Name:      csi-azurefile-vsc
Status:
  Bound Volume Snapshot Content Name:  snapcontent-c359a38f-35c1-4fb1-9da9-2c06d35ca0f4
  Ready To Use:                        false
Events:                                <none>

Resize a persistent volume

You can request a larger volume for a PVC. Edit the PVC object, and specify a larger size. This change triggers the expansion of the underlying volume that backs the PV.

In AKS, the built-in azurefile-csi storage class already supports expansion, so use the PVC created earlier with this storage class. The PVC requested a 100 GiB file share. We can confirm that by running:

kubectl exec -it nginx-azurefile -- df -h /mnt/azurefile

The output of the command resembles the following example:

Filesystem                                                                                Size  Used Avail Use% Mounted on
//f149b5a219bd34caeb07de9.file.core.chinacloudapi.cn/pvc-5e5d9980-da38-492b-8581-17e3cad01770  100G  128K  100G   1% /mnt/azurefile

Expand the PVC by increasing the spec.resources.requests.storage field:

kubectl patch pvc pvc-azurefile --type merge --patch '{"spec": {"resources": {"requests": {"storage": "200Gi"}}}}'

The output of the command resembles the following example:

persistentvolumeclaim/pvc-azurefile patched

Verify that both the PVC and the file system inside the pod show the new size:

kubectl get pvc pvc-azurefile
NAME            STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS    AGE
pvc-azurefile   Bound    pvc-5e5d9980-da38-492b-8581-17e3cad01770   200Gi      RWX            azurefile-csi   64m

kubectl exec -it nginx-azurefile -- df -h /mnt/azurefile
Filesystem                                                                                Size  Used Avail Use% Mounted on
//f149b5a219bd34caeb07de9.file.core.chinacloudapi.cn/pvc-5e5d9980-da38-492b-8581-17e3cad01770  200G  128K  200G   1% /mnt/azurefile

Use a persistent volume with private Azure Files storage (private endpoint)

If your Azure Files resources are protected with a private endpoint, you must create your own storage class. Make sure that you've configured your DNS settings to resolve the private endpoint IP address to the FQDN of the connection string. Customize the following parameters:

• resourceGroup: The resource group where the storage account is deployed.
• storageAccount: The storage account name.
• server: The FQDN of the storage account's private endpoint.

Create a file named private-azure-file-sc.yaml, and then paste the following example manifest in the file. Replace the values for <resourceGroup> and <storageAccountName>.

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: private-azurefile-csi
provisioner: file.csi.azure.com
allowVolumeExpansion: true
parameters:
  resourceGroup: <resourceGroup>
  storageAccount: <storageAccountName>
  server: <storageAccountName>.file.core.chinacloudapi.cn
reclaimPolicy: Delete
volumeBindingMode: Immediate
mountOptions:
  - dir_mode=0777
  - file_mode=0777
  - uid=0
  - gid=0
  - mfsymlinks
  - cache=strict  # https://linux.die.net/man/8/mount.cifs
  - nosharesock  # reduce probability of reconnect race
  - actimeo=30  # reduce latency for metadata-heavy workload

Create the storage class by using the kubectl apply command:

kubectl apply -f private-azure-file-sc.yaml

The output of the command resembles the following example:

storageclass.storage.k8s.io/private-azurefile-csi created

Create a file named private-pvc.yaml, and then paste the following example manifest in the file:

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: private-azurefile-pvc
spec:
  accessModes:
    - ReadWriteMany
  storageClassName: private-azurefile-csi
  resources:
    requests:
      storage: 100Gi

Create the PVC by using the kubectl apply command:

kubectl apply -f private-pvc.yaml

NFS file shares

Azure Files supports the NFS v4.1 protocol. NFS version 4.1 support for Azure Files provides you with a fully managed NFS file system as a service built on a highly available and highly durable distributed resilient storage platform.

This option is optimized for random access workloads with in-place data updates and provides full POSIX file system support. This section shows you how to use NFS shares with the Azure File CSI driver on an AKS cluster.

Prerequisites

• Your AKS cluster Control plane identity (that is, your AKS cluster name) is added to the Contributor role on the VNet and NetworkSecurityGroup.
• Your AKS cluster's service principal or managed service identity (MSI) must be added to the Contributor role to the storage account.

Optimizing read and write size options

This section provides information about how to approach performance tuning NFS with the Azure Files CSI driver with the rsize and wsize options. The rsize and wsize options set the maximum transfer size of an NFS operation. If rsize or wsize are not specified on mount, the client and server negotiate the largest size supported by the two. Currently, both Azure NetApp Files and modern Linux distributions support read and write sizes as large as 1,048,576 Bytes (1 MiB).

Optimal performance is based on efficient client-server communication. Increasing or decreasing the mount read and write option size values can improve NFS performance. The default size of the read/write packets transferred between client and server are 8 KB for NFS version 2, and 32 KB for NFS version 3 and 4. These defaults may be too large or too small. Reducing the rsize and wsize might improve NFS performance in a congested network by sending smaller packets for each NFS-read reply and write request. However, this can increase the number of packets needed to send data across the network, increasing total network traffic and CPU utilization on the client and server.

It's important that you perform testing to find an rsize and wsize that sustains efficent packet transfer, where it doesn't decrease throughput and increase latency.

For example, to configure a maximum rsize and wsize of 256-KiB, configure the mountOptions in the storage class as follows:

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: azurefile-csi-nfs
provisioner: file.csi.azure.com
allowVolumeExpansion: true
parameters:
  protocol: nfs
mountOptions:
  - nconnect=4
  - noresvport
  - actimeo=30
  - rsize=262144
  - wsize=262144

Create NFS file share storage class

Create a file named nfs-sc.yaml and copy the manifest below. For a list of supported mountOptions, see NFS mount options.

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: azurefile-csi-nfs
provisioner: file.csi.azure.com
allowVolumeExpansion: true
parameters:
  protocol: nfs
mountOptions:
  - nconnect=4
  - noresvport
  - actimeo=30

After editing and saving the file, create the storage class with the kubectl apply command:

kubectl apply -f nfs-sc.yaml

The output of the command resembles the following example:

storageclass.storage.k8s.io/azurefile-csi-nfs created

Create a deployment with an NFS-backed file share

You can deploy an example stateful set that saves timestamps into a file data.txt with the kubectl apply command:

kubectl apply -f

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: statefulset-azurefile
  labels:
    app: nginx
spec:
  podManagementPolicy: Parallel  # default is OrderedReady
  serviceName: statefulset-azurefile
  replicas: 1
  template:
    metadata:
      labels:
        app: nginx
    spec:
      nodeSelector:
        "kubernetes.io/os": linux
      containers:
        - name: statefulset-azurefile
          image: mcr.azk8s.cn/oss/nginx/nginx:1.19.5
          command:
            - "/bin/bash"
            - "-c"
            - set -euo pipefail; while true; do echo $(date) >> /mnt/azurefile/outfile; sleep 1; done
          volumeMounts:
            - name: persistent-storage
              mountPath: /mnt/azurefile
  updateStrategy:
    type: RollingUpdate
  selector:
    matchLabels:
      app: nginx
  volumeClaimTemplates:
    - metadata:
        name: persistent-storage
      spec:
        storageClassName: azurefile-csi-nfs
        accessModes: ["ReadWriteMany"]
        resources:
          requests:
            storage: 100Gi

The output of the command resembles the following example:

statefulset.apps/statefulset-azurefile created

Validate the contents of the volume by running the following command:

kubectl exec -it statefulset-azurefile-0 -- df -h

The output of the command resembles the following example:

Filesystem      Size  Used Avail Use% Mounted on
...
/dev/sda1                                                                                 29G   11G   19G  37% /etc/hosts
accountname.file.core.chinacloudapi.cn:/accountname/pvc-fa72ec43-ae64-42e4-a8a2-556606f5da38  100G     0  100G   0% /mnt/azurefile
...

Windows containers

The Azure Files CSI driver also supports Windows nodes and containers. To use Windows containers, follow the Windows containers quickstart to add a Windows node pool.

After you have a Windows node pool, use the built-in storage classes like azurefile-csi or create a custom one. You can deploy an example Windows-based stateful set that saves timestamps into a file data.txt by running the kubectl apply command:

kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/azurefile-csi-driver/master/deploy/example/windows/statefulset.yaml

The output of the command resembles the following example:

statefulset.apps/busybox-azurefile created

Validate the contents of the volume by running the following kubectl exec command:

kubectl exec -it busybox-azurefile-0 -- cat c:\\mnt\\azurefile\\data.txt # on Linux/MacOS Bash
kubectl exec -it busybox-azurefile-0 -- cat c:\mnt\azurefile\data.txt # on Windows Powershell/CMD

The output of the commands resembles the following example:

2020-08-27 22:11:01Z
2020-08-27 22:11:02Z
2020-08-27 22:11:04Z
(...)

Next steps

• To learn how to use CSI driver for Azure Disks, see Use Azure Disks with CSI driver.
• To learn how to use CSI driver for Azure Blob storage, see Use Azure Blob storage with CSI driver.
• For more about storage best practices, see Best practices for storage and backups in Azure Kubernetes Service.