Deploy Azure Local rack aware cluster using Azure Resource Manager deployment template (Preview)

Applies to: Azure Local version 2510 and later

This article describes how to use an Azure Resource Manager (ARM) template in the Azure portal to deploy a rack aware cluster.

Prerequisites

• Completion of Register your machines with Azure Arc and assign deployment permissions. Make sure that:

  • All machines are running the same version of OS.
  • All the machines have the same network adapter configuration.

• Make sure to select the create-cluster-rac-enabled template for deployment.

Step 1: Prepare Azure resources

Follow these steps to prepare the Azure resources you need for deployment:

Get the object ID for Azure Local Resource Provider

This object ID for the Azure Local Resource Provider (RP) is unique per Azure tenant.

1. In the Azure portal, search for and go to Microsoft Entra ID.

2. Go to the Overview tab and search for Microsoft.AzureStackHCI Resource Provider.

   

3. Select the Service Principal Name that is listed and copy the Object ID.

   

   Alternatively, you can use PowerShell to get the object ID of the Azure Local RP service principal. Run the following command in PowerShell:

   Get-AzADServicePrincipal -DisplayName "Microsoft.AzureStackHCI Resource Provider"
   

Step 2: Deploy using ARM template

An ARM template is a JSON file where you define what you want to deploy to Azure. It creates and assigns all the resource permissions required for deployment.

Once all prerequisite and preparation steps are complete, you're ready to deploy using a validated ARM template and its corresponding parameters JSON file. This file contains all required values, including those generated previously.

ARM template deployment involves two modes:

• Validate: Confirms that all parameters are correctly configured and validates your system's readiness to deploy.
• Deploy: Performs the actual deployment after successful validation.

Step 2.1: Deploy the template in Validate mode

This step ensures that all parameters are configured correctly and validates your system's readiness to deploy.

1. In the Azure portal, go to Home and then select + Create a resource.

2. Under Template deployment (deploy using custom templates), select Create.

   

3. On the Custom deployment page, proceed through the tabs described in the following sections.

Select a template tab

Use the Select a template tab to choose the template for your deployment.

1. On the Select a template tab, under the Start with a quickstart template or template spec section, select the Quickstart template option.

2. From the Quickstart template (disclaimer) dropdown, select the create-cluster-rac-enabled template.

3. Select the Select template button to continue to the Basics tab.

   

Basics tab

Use the Basics tab to provide the essential information to initiate the deployment.

This section describes how to configure the Basics tab using the following sample rack aware cluster configuration:

You deploy a rack aware cluster with four machines, two in each rack:

• node1 and node2 are physically located in the same rack (Zone1).
• node3 and node4 are located in a different rack (Zone2).

For an example of a parameter JSON file that shows the format of various inputs, such as ArcNodeResourceId, see azuredeploy.parameters.json. For detailed descriptions of the parameters defined in this file, see ARM template parameters reference.

1. On the Basics tab, select the required parameters from the dropdown list, or select Edit parameters to modify them manually.

   

2. Configure all required parameters. Configure the following extra parameters required for rack aware cluster deployment. You can use the sample JSON snippets to deploy the cluster described in the example configuration.

   • Arc node resource IDs.

     "arcNodeResourceIds": {
     "value": [
     "/subscriptions/<SubscriptionID>/resourceGroups/<ResourceGroupName>/providers/Microsoft.HybridCompute/machines/node1",
     "/subscriptions/<SubscriptionID>/resourceGroups/<ResourceGroupName>/providers/Microsoft.HybridCompute/machines/node2",
     "/subscriptions/<SubscriptionID>/resourceGroups/<ResourceGroupName>/providers/Microsoft.HybridCompute/machines/node3",
     "/subscriptions/<SubscriptionID>/resourceGroups/<ResourceGroupName>/providers/Microsoft.HybridCompute/machines/node4"
     ]
     }
     

   • Cluster pattern and local availability zones. Verify that the clusterPattern parameter is RackAware. Ensure that node1 and node2 are physically located in Zone1 (same rack), and node3 and node4 are in Zone2 (different rack).

     "clusterPattern": {
     "value": "RackAware"
     },
     "localAvailabilityZones": {
         "value": [
         {
             "localAvailabilityZoneName": "Zone1",
             "nodes": ["node1","node2"]  
         },
         {
             "localAvailabilityZoneName": "Zone2",
             "nodes": ["node3","node4"]
         }
         ]
     },
     

   • Cloud witness configuration. Rack aware cluster requires cloud witness. Enter the name of the cloud witness, which is created during the deployment process.

     "witnessType": {
         "value": "Cloud"
     },
     "clusterWitnessStorageAccountName": {
         "value": "yourcloudwitness"
     },
     

   • Network intent configuration. The storage network intent must be a dedicated network intent. VLAN IDs 711 and 712 are defaults and can be customized for your environment.

     "networkingType": {
         "value": "switchedMultiserverDeployment"
     },
     "networkingPattern": {
         "value" : "convergedManagementCompute"
     },
     "intentList" : {
         "value": [
             {
                 "name": "ManagementCompute",
                 "trafficType": [
                     "Management",
                     "Compute"
                 ],
                 "adapter": [
                     "ethernet",
                     "ethernet 2"
                 ],
                 "overridevirtualswitchConfiguration": false,
                 "virtualswitchConfigurationoverrides": {
                     "enableIov": "",
                     "loadBalancingAlgorithm": ""
                 },
                 "overrideQosPolicy": false,
                 "qosPolicyoverrides": {
                 "priorityvalue8021Action_SMB": "",
                 "priorityvalues8021Action_Cluster": "",
                 "bandwidthPercentage_SMB": ""
                 },
                 "overrideAdapterProperty": false,
                 "adapterPropertyoverrides": {
                     "jumboPacket": "",
                     "networkDirect": "",
                     "networkDirectTechnology": ""
                 }
             },
             "name": "Storage",
             "trafficType": [
                     "Storage"
                     ],
                 "adapter": [
                     "ethernet 3",
                     "ethernet 4"
                 ],
                 "overridevirtualswitchConfiguration": false,
                 "virtualswitchConfigurationoverrides": {
                     "enableIov": "",
                     "loadBalancingAlgorithm": ""
                 },
                 "overrideQosPolicy": false,
                 "qosPolicyoverrides": {
                 "priorityvalue8021Action_SMB": "",
                 "priorityvalues8021Action_Cluster": "",
                 "bandwidthPercentage_SMB": ""
                 },
                 "overrideAdapterProperty": false,
                 "adapterPropertyoverrides": {
                     "jumboPacket": "",
                     "networkDirect": "",
                     "networkDirectTechnology": ""
                 }
             ]
     },
     "storageNetworkList": {
         "value": [
             {
                 "name": "Storage1Network",
                 "networkAdapterName": "ethernet 3",
                 "vlanId": "711"
             },
             {
                 "name": "Storage2Network",
                 "networkAdapterName": "ethernet 4",
                 "vlanId": "712"
             }
         ]
     },            
     

3. After configuring all parameters, select Save to save the parameters file.

4. Select the appropriate resource group for your environment.

5. Confirm that Deployment Mode is set to Validate.

6. Select Review + create to continue.

   

Review + create tab

Use the Review + create tab to review your deployment settings and accept the legal terms before creating the resources.

1. On the Review + Create tab, review the deployment summary and legal terms.

2. Select Create to begin validation. This action creates the remaining prerequisite resources and validates the deployment. Validation typically takes about 15 minutes to deploy one to two machines and longer for bigger deployments. Monitor the validation progress.

   

3. After validation completes, continue to the Deploy phase to provision the full environment.

Step 2.2: Deploy the template in Deploy mode

After successful validation, you're ready to proceed with the actual deployment using the validated ARM template and its corresponding parameters JSON file.

1. Once validation is complete, select Redeploy.

   

2. On the Custom deployment screen, select Edit parameters. Load up the previously saved parameters and select Save.

3. Verify that all fields for the ARM template are filled in by the parameters JSON file.

4. Select the appropriate resource group for your environment.

5. Confirm that Deployment Mode is set to Deploy.

   

6. Select Review + create.

7. Select Create to begin deployment. The deployment uses the existing prerequisite resources created during the Deploy the template in Validate mode step.

   The Deployment screen cycles on the cluster resource during deployment.

   Once deployment initiates, there's a limited Environment Checker run, a full Environment Checker run, and cloud deployment starts.

   After a few minutes, you can monitor deployment progress in the Azure portal.

   

Monitor deployment

1. In a new browser window, navigate to the resource group for your environment. Select the cluster resource.

2. Select Deployments.

3. Refresh and watch the deployment progress from the first machine (also known as the seed machine and is the first machine where you deployed the cluster). Deployment takes between 2.5 and 3 hours. Several steps take 40-50 minutes or more.

4. The step in deployment that takes the longest is Deploy Moc and ARB Stack. This step takes 40-45 minutes.

   Once complete, the task at the top updates with status and end time.

ARM template parameters reference

The following table describes the parameters that you define in the ARM template's parameters file:

The following table describes parameters that are specific to a rack aware cluster:

Next steps

• After deployment is complete, continue to Complete post-deployment tasks for rack aware cluster.