Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
If multiple pods need concurrent access to the same storage volume, you can use Azure Blob storage to connect using blobfuse or Network File System (NFS).
This article shows you how to dynamically and statically create Azure Blob storage containers for use by multiple pods in an Azure Kubernetes Service (AKS) cluster.
Prerequisites
The Azure Blob storage CSI driver enabled on your AKS cluster.
To support an Azure DataLake Gen2 storage account when using blobfuse mount, you need to do the following:
- To create an ADLS account using the driver in dynamic provisioning, specify
isHnsEnabled: "true"in the storage class parameters. - To enable blobfuse access to an ADLS account in static provisioning, specify the mount option
--use-adls=truein the persistent volume. - If you're going to enable a storage account with Hierarchical Namespace, existing persistent volumes should be remounted with
--use-adls=truemount option.
- To create an ADLS account using the driver in dynamic provisioning, specify
By default, the blobfuse cache is located in the
/mntdirectory. If the VM SKU provides a temporary disk, the/mntdirectory is mounted on the temporary disk. However, if the VM SKU doesn't provide a temporary disk, the/mntdirectory is mounted on the OS disk, you can set--tmp-path=mount option to specify a different cache directory.
Use built-in storage classes to create dynamic PVs with Azure Blob storage
A storage class is used to define how an Azure Blob storage container is created. A storage account is automatically created in the node resource group for use with the storage class to hold the Azure Blob storage container. When you use storage CSI drivers on AKS, there are two extra built-in StorageClasses that use the Azure Blob storage CSI driver.
The reclaim policy on both storage classes ensures that the underlying Azure Blob storage is deleted when the respective PV is deleted. The storage classes also configure the container to be expandable by default, as the set allowVolumeExpansion parameter is set to true.
Note
Shrinking persistent volumes isn't supported.
You can select one of the following Azure storage redundancy SKUs for the skuname parameter in the storage class definition:
- Standard_LRS: Standard locally redundant storage
- Premium_LRS: Premium locally redundant storage
- Standard_ZRS: Standard zone redundant storage
- Premium_ZRS: Premium zone redundant storage
- Standard_GRS: Standard geo-redundant storage
- Standard_RAGRS: Standard read-access geo-redundant storage
Create custom storage classes for dynamic PVs with Azure Blob storage
The default storage classes are suitable for most scenarios. In some cases, you might want to have your own storage class customized with your own parameters. In this section, we provide two examples: one using NFS protocol and one using blobfuse.
Custom storage class example using NFS protocol
The manifest in this example mounts a blob storage container using the NFS protocol. You can use it to add the tags parameter.
Create a file named
blob-nfs-sc.yamland paste in the following example manifest:apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: azureblob-nfs-premium provisioner: blob.csi.azure.com parameters: protocol: nfs tags: environment=Development volumeBindingMode: Immediate allowVolumeExpansion: true mountOptions: - nconnect=4Create the storage class using the
kubectl applycommand:kubectl apply -f blob-nfs-sc.yamlYour output should resemble the following example output:
storageclass.storage.k8s.io/blob-nfs-premium created
Custom storage class example using blobfuse
The manifest in this example uses blobfuse and mounts a Blob storage container. You can use it to update the skuName parameter.
Create a file named
blobfuse-sc.yamland paste in the following example manifest:apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: azureblob-fuse-premium provisioner: blob.csi.azure.com parameters: skuName: Standard_GRS # available values: Standard_LRS, Premium_LRS, Standard_GRS, Standard_RAGRS reclaimPolicy: Delete volumeBindingMode: Immediate allowVolumeExpansion: true mountOptions: - -o allow_other - --file-cache-timeout-in-seconds=120 - --use-attr-cache=true - --cancel-list-on-mount-seconds=10 # prevent billing charges on mounting - -o attr_timeout=120 - -o entry_timeout=120 - -o negative_timeout=120 - --log-level=LOG_WARNING # LOG_WARNING, LOG_INFO, LOG_DEBUG - --cache-size-mb=1000 # Default will be 80% of available memory, eviction will happen beyond that.Create the storage class using the
kubectl applycommand:kubectl apply -f blobfuse-sc.yamlYour output should resemble the following example output:
storageclass.storage.k8s.io/blob-fuse-premium created
Storage class parameters for dynamic PVs with Azure Blob storage
The following table includes parameters you can use to define a custom storage class for your persistent volume claims (PVCs) with Azure Blob storage:
| Name | Meaning | Available values | Required | Default value |
|---|---|---|---|---|
skuName |
Specify an Azure storage account type (alias: storageAccountType). |
Standard_LRS, Premium_LRS, Standard_GRS, Standard_RAGRS |
No | Standard_LRS |
location |
Specify an Azure location. | eastus |
No | If empty, driver uses the same location name as current cluster. |
resourceGroup |
Specify an Azure resource group name. | myResourceGroup | No | If empty, driver uses the same resource group name as current cluster. |
storageAccount |
Specify an Azure storage account name. | storageAccountName | No | When a specific storage account name isn't provided, the driver looks for a suitable storage account that matches the account settings within the same resource group. If it fails to find a matching storage account, it creates a new one. However, if a storage account name is specified, the storage account must already exist. |
networkEndpointType |
Specify network endpoint type for the storage account created by driver. If privateEndpoint is specified, a private endpoint is created for the storage account. For other cases, a service endpoint will be created for NFS protocol. | privateEndpoint |
No | For an AKS cluster, add the AKS cluster name to the Contributor role in the resource group hosting the VNet. |
protocol |
Specify blobfuse mount or NFSv3 mount. | fuse, nfs |
No | fuse |
containerName |
Specify the existing container (directory) name. | container | No | If empty, driver creates a new container name, starting with pvc-fuse for blobfuse or pvc-nfs for NFS v3. |
containerNamePrefix |
Specify Azure storage directory prefix created by driver. | my | Can only contain lowercase letters, numbers, hyphens, and length should be fewer than 21 characters. | No |
server |
Specify Azure storage account domain name. | Existing storage account DNS domain name, for example <storage-account>.blob.core.windows.net. |
No | If empty, driver uses default <storage-account>.blob.core.windows.net or other sovereign cloud storage account DNS domain name. |
allowBlobPublicAccess |
Allow or disallow public access to all blobs or containers for storage account created by driver. | true,false |
No | false |
storageEndpointSuffix |
Specify Azure storage endpoint suffix. | core.windows.net |
No | If empty, driver uses default storage endpoint suffix according to cloud environment. |
tags |
Tags would be created in new storage account. | Tag format: 'foo=aaa,bar=bbb' | No | "" |
matchTags |
Match tags when driver tries to find a suitable storage account. | true,false |
No | false |
| --- | The following parameters are only for blobfuse | --- | --- | --- |
subscriptionID |
Specify Azure subscription ID where blob storage directory will be created. | Azure subscription ID | No | If not empty, resourceGroup must be provided. |
storeAccountKey |
Specify store account key to Kubernetes secret. Note: false means driver uses kubelet identity to get account key. |
true,false |
No | true |
secretName |
Specify secret name to store account key. | No | ||
secretNamespace |
Specify the namespace of secret to store account key. | default,kube-system, etc. |
No | PVC namespace |
isHnsEnabled |
Enable Hierarchical namespace for Azure Data Lake storage account. |
true,false |
No | false |
| --- | The following parameters are only for NFS protocol | --- | --- | --- |
mountPermissions |
Specify mounted folder permissions. | The default is 0777. If set to 0, driver won't perform chmod after mount. |
No | 0777 |
Note
If the storage account is created by the driver, then you only need to specify networkEndpointType: privateEndpoint parameter in storage class. The CSI driver creates the private endpoint and private DNS zone (named privatelink.blob.core.windows.net) together with the account. If you bring your own storage account, then you need to create the private endpoint for the storage account. If you're using Azure Blob storage in a network isolated cluster, you must create a custom storage class with "networkEndpointType: privateEndpoint". You can use the following example manifest as a reference:
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: blob-fuse
provisioner: blob.csi.azure.com
parameters:
skuName: Premium_LRS # available values: Standard_LRS, Premium_LRS, Standard_GRS, Standard_RAGRS, Standard_ZRS, Premium_ZRS
protocol: fuse2
networkEndpointType: privateEndpoint
reclaimPolicy: Delete
volumeBindingMode: Immediate
allowVolumeExpansion: true
mountOptions:
- -o allow_other
- --file-cache-timeout-in-seconds=120
- --use-attr-cache=true
- --cancel-list-on-mount-seconds=10 # prevent billing charges on mounting
- -o attr_timeout=120
- -o entry_timeout=120
- -o negative_timeout=120
- --log-level=LOG_WARNING # LOG_WARNING, LOG_INFO, LOG_DEBUG
- --cache-size-mb=1000 # Default will be 80% of available memory, eviction will happen beyond that.
Create a PVC with Azure Blob storage
A PVC uses the storage class object to dynamically provision an Azure Blob storage. You can use the example YAML manifest in this section to create a PVC that's 5 GB in size with ReadWriteMany access. For more information on access modes, see Kubernetes PV access modes.
Create a file named
blob-nfs-pvc.yamland paste in the following YAML manifest:apiVersion: v1 kind: PersistentVolumeClaim metadata: name: azure-blob-storage spec: accessModes: - ReadWriteMany storageClassName: azureblob-nfs-premium resources: requests: storage: 5GiCreate the PVC using the
kubectl createcommand:kubectl create -f blob-nfs-pvc.yamlView the status of the PVC with the [
kubectl get pvc][kubectl-get-pvc] command:kubectl get pvc azure-blob-storageYour output should resemble the following example output, which shows that the PVC is in a
Boundstate:NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE azure-blob-storage Bound pvc-aaaaaaaa-0000-1111-2222-bbbbbbbbbbbb 5Gi RWX azureblob-nfs-premium 92m
Use a PVC in a pod to mount Azure Blob storage
The following YAML creates a pod that uses the persistent volume claim azure-blob-storage to mount the Azure Blob storage at the `/mnt/blob' path.
Create a file named
blob-nfs-pvand paste in the following YAML manifest. Make sure theclaimNamematches the PVC created in the previous step.kind: Pod apiVersion: v1 metadata: name: mypod spec: containers: - name: mypod image: mcr.microsoft.com/oss/nginx/nginx:1.17.3-alpine resources: requests: cpu: 100m memory: 128Mi limits: cpu: 250m memory: 256Mi volumeMounts: - mountPath: "/mnt/blob" name: volume readOnly: false volumes: - name: volume persistentVolumeClaim: claimName: azure-blob-storageCreate the pod using the
kubectl applycommand:kubectl apply -f blob-nfs-pv.yamlOnce the pod is successfully running, create a new file named
test.txtusing the following command:kubectl exec mypod -- touch /mnt/blob/test.txtValidate the disk is correctly mounted using the following command to list the files in the mounted directory:
kubectl exec mypod -- ls /mnt/blobYour output should resemble the following example output, which shows the
test.txtfile you created in the mounted Azure Blob storage:test.txt
Use a StatefulSet to manage the lifecycle of a volume with Azure Blob storage
To have a storage volume persist for your workload, you can use a StatefulSet. This makes it easier to match existing volumes to new pods that replace any that have failed. The following examples demonstrate how to set up a StatefulSet for Blob storage using the NFS protocol or Blobfuse.
Note
IF you're using NFS protocol, your AKS cluster control plane identity (that is, your AKS cluster name) needs to be added to the Contributor role on the VNet and network security group.
Create a file named
azure-blob-nfs-ss.yamland paste in the following YAML manifest:apiVersion: apps/v1 kind: StatefulSet metadata: name: statefulset-blob-nfs labels: app: nginx spec: serviceName: statefulset-blob-nfs replicas: 1 template: metadata: labels: app: nginx spec: nodeSelector: "kubernetes.io/os": linux containers: - name: statefulset-blob-nfs image: mcr.microsoft.com/azurelinux/base/nginx:1.25 volumeMounts: - name: persistent-storage mountPath: /mnt/blob updateStrategy: type: RollingUpdate selector: matchLabels: app: nginx volumeClaimTemplates: - metadata: name: persistent-storage spec: storageClassName: azureblob-nfs-premium accessModes: ["ReadWriteMany"] resources: requests: storage: 100GiCreate the StatefulSet using the
kubectl createcommand:kubectl create -f azure-blob-nfs-ss.yaml
Create a static PV with Azure Blob storage
The following sections provide instructions for creating a static PV with Azure Blob storage. A static PV is a persistent volume that an administrator creates manually. This PV is available for use by pods in the cluster. To use a static PV, you create a PVC that references the PV, and then create a pod that references the PVC.
Storage class parameters for static PVs with Azure Blob storage
The following table includes parameters you can use to define a custom storage class for your static PVCs with Azure Blob storage:
| Name | Meaning | Available values | Required | Default value |
|---|---|---|---|---|
volumeHandle |
Specify a value the driver can use to uniquely identify the storage blob container in the cluster. | A recommended way to produce a unique value is to combine the globally unique storage account name and container name: {account-name}_{container-name}.Note: The #, / characters are reserved for internal use and can't be used in a volume handle. |
Yes | |
volumeAttributes.resourceGroup |
Specify Azure resource group name. | myResourceGroup | No | If empty, driver uses the same resource group name as current cluster. |
volumeAttributes.storageAccount |
Specify an existing Azure storage account name. | storageAccountName | Yes | |
volumeAttributes.containerName |
Specify existing container name. | container | Yes | |
volumeAttributes.protocol |
Specify blobfuse mount or NFS v3 mount. | fuse, nfs |
No | fuse |
| --- | The following parameters are only for blobfuse | --- | --- | --- |
volumeAttributes.secretName |
Secret name that stores storage account name and key (only applies for SMB). | No | ||
volumeAttributes.secretNamespace |
Specify namespace of secret to store account key. | default |
No | PVC namespace |
nodeStageSecretRef.name |
Specify secret name that stores one of the following:azurestorageaccountkeyazurestorageaccountsastokenmsisecretazurestoragespnclientsecret. |
No | Existing Kubernetes secret name | |
nodeStageSecretRef.namespace |
Specify the namespace of secret. | Kubernetes namespace | Yes | |
| --- | The following parameters are only for NFS protocol | --- | --- | --- |
volumeAttributes.mountPermissions |
Specify mounted folder permissions. | 0777 |
No | |
| --- | The following parameters are only for NFS VNet setting | --- | --- | --- |
vnetResourceGroup |
Specify VNet resource group hosting virtual network. | myResourceGroup | No | If empty, driver uses the vnetResourceGroup value specified in the Azure cloud config file. |
vnetName |
Specify the virtual network name. | aksVNet | No | If empty, driver uses the vnetName value specified in the Azure cloud config file. |
subnetName |
Specify the existing subnet name of the agent node. | aksSubnet | No | if empty, driver will update all the subnets under the cluster virtual network. |
| --- | The following parameters are only for feature: blobfuse Managed Identity and Service Principal Name authentication |
--- | --- | --- |
volumeAttributes.AzureStorageAuthType |
Specify the authentication type. | Key, SAS, MSI, SPN |
No | Key |
volumeAttributes.AzureStorageIdentityClientID |
Specify the Identity Client ID. | No | ||
volumeAttributes.AzureStorageIdentityResourceID |
Specify the Identity Resource ID. | No | ||
volumeAttributes.MSIEndpoint |
Specify the MSI endpoint. | No | ||
volumeAttributes.AzureStorageSPNClientID |
Specify the Azure Service Principal Name (SPN) Client ID. | No | ||
volumeAttributes.AzureStorageSPNTenantID |
Specify the Azure SPN Tenant ID. | No | ||
volumeAttributes.AzureStorageAADEndpoint |
Specify the Microsoft Entra endpoint. | No | ||
| --- | The following parameters are only for feature: blobfuse read account key or SAS token from key vault | --- | --- | --- |
volumeAttributes.keyVaultURL |
Specify Azure Key Vault DNS name. | {vault-name}.vault.azure.net | No | |
volumeAttributes.keyVaultSecretName |
Specify Azure Key Vault secret name. | Existing Azure Key Vault secret name. | No | |
volumeAttributes.keyVaultSecretVersion |
Azure Key Vault secret version. | Existing version | No | If empty, driver uses current version. |
Create a Blob storage container
When you create an Azure Blob storage resource for use with AKS, you can create the resource in the node resource group. This approach allows the AKS cluster to access and manage the blob storage resource.
Get the node resource group name of your AKS cluster using the
az aks showcommand with the--query nodeResourceGroupparameter.az aks show --resource-group myResourceGroup --name myAKSCluster --query nodeResourceGroup -o tsvThe output of the command resembles the following example:
MC_myResourceGroup_myAKSCluster_eastusCreate a container for storing blobs following the steps in the Manage blob storage to authorize access and then create the container.
Mount volume
In this section, you mount the persistent volume using the NFS protocol or Blobfuse.
Mounting Blob storage using the NFS v3 protocol doesn't authenticate using an account key. Your AKS cluster needs to reside in the same or peered virtual network as the agent node. The only way to secure the data in your storage account is by using a virtual network and other network security settings. For more information on how to set up NFS access to your storage account, see Mount Blob Storage by using the Network File System (NFS) 3.0 protocol.
The following example demonstrates how to mount a Blob storage container as a persistent volume using the NFS protocol.
Create a file named
pv-blob-nfs.yamland paste in the following YAML. UnderstorageClass, updateresourceGroup,storageAccount, andcontainerName.Note
volumeHandlevalue should be a unique volumeID for every identical storage blob container in the cluster. The character#and/are reserved for internal use and can't be used.apiVersion: v1 kind: PersistentVolume metadata: annotations: pv.kubernetes.io/provisioned-by: blob.csi.azure.com name: pv-blob spec: capacity: storage: 1Pi accessModes: - ReadWriteMany persistentVolumeReclaimPolicy: Retain # If set as "Delete" container would be removed after pvc deletion storageClassName: azureblob-nfs-premium mountOptions: - nconnect=4 csi: driver: blob.csi.azure.com # make sure volumeid is unique for every identical storage blob container in the cluster # character `#` and `/` are reserved for internal use and cannot be used in volumehandle volumeHandle: account-name_container-name volumeAttributes: resourceGroup: resourceGroupName storageAccount: storageAccountName containerName: containerName protocol: nfsNote
While the Kubernetes API capacity attribute is mandatory, this value isn't used by the Azure Blob storage CSI driver because you can flexibly write data until you reach your storage account's capacity limit. The value of the
capacityattribute is used only for size matching between PVs and PVCs. We recommend using a fictitious high value. The pod sees a mounted volume with a fictitious size of 5 Petabytes.Create the PV using the
kubectl createcommand:kubectl create -f pv-blob-nfs.yamlCreate a file named
pvc-blob-nfs.yamland paste in the following YAML. UndervolumeName, update the value to match the name of the PV created in the previous step.kind: PersistentVolumeClaim apiVersion: v1 metadata: name: pvc-blob spec: accessModes: - ReadWriteMany resources: requests: storage: 10Gi volumeName: pv-blob storageClassName: azureblob-nfs-premiumCreate the PVC using the
kubectl createcommand:kubectl create -f pvc-blob-nfs.yaml
Use the persistent volume
The following YAML creates a pod that uses the PV or PVC named pvc-blob created earlier to mount the Azure Blob storage at the /mnt/blob path.
Create a file named
nginx-pod-blob.yamland paste in the following YAML manifest. Make sure theclaimNamematches the PVC created in the previous step when creating a PV for NFS or Blobfuse.kind: Pod apiVersion: v1 metadata: name: nginx-blob spec: nodeSelector: "kubernetes.io/os": linux containers: - image: mcr.microsoft.com/oss/nginx/nginx:1.17.3-alpine name: nginx-blob volumeMounts: - name: blob01 mountPath: "/mnt/blob" readOnly: false volumes: - name: blob01 persistentVolumeClaim: claimName: pvc-blobCreate the pod and mount the PVC using the
kubectl createcommand:kubectl create -f nginx-pod-blob.yamlCreate an interactive shell session with the pod to verify the Blob storage is mounted correctly using the following
kubectl execcommand:kubectl exec -it nginx-blob -- df -hYour output should resemble the following example output, which shows the Blob storage is mounted at the
/mnt/blobpath:Filesystem Size Used Avail Use% Mounted on ... blobfuse 14G 41M 13G 1% /mnt/blob ...