Deploy your application on Virtual Machine Scale Sets

To run applications on virtual machine (VM) instances in a scale set, you first need to install the application components and required files. This article introduces ways to build a custom VM image for instances in a scale set, or automatically run install scripts on existing VM instances. You also learn how to manage application or OS updates across a scale set.

Azure VM Apps provide a streamlined, scalable and secure way to package, manage, and deploy applications across Azure Virtual Machines (VMs), Virtual Machine Scale Sets (VMSS). VM Apps is the best suited method for deploying tools, agents, AI Apps, security components, and line-of-business applications across Azure.

• Centralized and Flexible Application Management:
  • Package Once, Deploy Anywhere: Package applications in ZIP, MSI, or EXE formats and manage them centrally in Azure Compute Gallery.
  • Version Control: Maintain multiple versions of applications, allowing teams to deploy the latest or a specific version as needed.
• Seamless Sharing and Access Control
  • Tenant-Wide Sharing: Share applications within teams or across your entire organization (tenant).
  • Integrated RBAC: Control publishing and deployment access using Azure Role-Based Access Control (RBAC).
• Reliable and Customizable Deployments
  • Individual App Control: Install, update, or delete applications independently—no need to rebuild VM images.
  • Customizable Operations: Define how install, update, and delete operations for applications, including reboot handling.
  • Built-In Failure Handling: Ensure resilient deployments by connecting VM application failure to VM failure.
• Scalable and Low-Latency Distribution
  • Global and Intra-Region Replication: Automatically replicate applications across and within regions to reduce latency and improve resiliency—no need for AzCopy or manual transfers.
  • Optimized for High-Scale Scenarios: Achieve low create latency even during large-scale deployments.
• Secure and Compliant by Design
  • Policy-Driven Enforcement: Use Azure Policy to enforce application presence and configuration across your fleet.
  • Secure Deployments: Avoid internet-based downloads and complex private link setups—ideal for locked-down environments.
• Broad Platform Support
  • VMs and Scale Sets: Deploy to individual VMs, flexible scale sets, or uniform scale sets with full support.
  • Block Blob Support: Efficiently handle large application packages using Azure Block Blobs for chunked uploads and background streaming.

• Overview of VM Apps
• Create and deploy VM Apps on Linux VM
• Create and deploy VM Apps on Windows VM

When you use one of the Azure platform images to create the instances in your scale set, no other software is installed or configured. You can automate the install of these components, however that adds to the time it takes to provision VM instances to your scale sets. If you apply many configuration changes to the VM instances, there's management overhead with those configuration scripts and tasks.

To reduce the configuration management and time to provision a VM, you can create a custom VM image that is ready to run your application as soon as an instance is provisioned in the scale set. For more information on how to create and use a custom VM image with a scale set, see the following tutorials:

• Azure CLI
• Azure PowerShell

The Custom Script Extension downloads and executes scripts on Azure VMs. This extension is useful for post deployment configuration, software installation, or any other configuration / management task. Scripts can be downloaded from Azure storage or GitHub, or provided to the Azure portal at extension run-time. For more information on how to install an app with a Custom Script Extension, see the following tutorials:

• Azure CLI
• Azure PowerShell
• Azure Resource Manager template

PowerShell Desired State Configuration (DSC) is a management platform to define the configuration of target machines. DSC configurations define what to install on a machine and how to configure the host. A Local Configuration Manager (LCM) engine runs on each target node that processes requested actions based on pushed configurations.

The PowerShell DSC extension lets you customize VM instances in a scale set with PowerShell. The following example:

• Instructs the VM instances to download a DSC package from GitHub - https://github.com/Azure-Samples/compute-automation-configurations/raw/master/dsc.zip
• Sets the extension to run an install script - configure-http.ps1
• Gets information about a scale set with Get-AzVmss
• Applies the extension to the VM instances with Update-AzVmss

The DSC extension is applied to the myScaleSet VM instances in the resource group named myResourceGroup. Enter your own names as follows:

# Define the script for your Desired Configuration to download and run
$dscConfig = @{
  "wmfVersion" = "latest";
  "configuration" = @{
    "url" = "https://github.com/Azure-Samples/compute-automation-configurations/raw/master/dsc.zip";
    "script" = "configure-http.ps1";
    "function" = "WebsiteTest";
  };
}

# Get information about the scale set
$vmss = Get-AzVmss `
                -ResourceGroupName "myResourceGroup" `
                -VMScaleSetName "myScaleSet"

# Add the Desired State Configuration extension to install IIS and configure basic website
$vmss = Add-AzVmssExtension `
    -VirtualMachineScaleSet $vmss `
    -Publisher Microsoft.Powershell `
    -Type DSC `
    -TypeHandlerVersion 2.24 `
    -Name "DSC" `
    -Setting $dscConfig

# Update the scale set and apply the Desired State Configuration extension to the VM instances
Update-AzVmss `
    -ResourceGroupName "myResourceGroup" `
    -Name "myScaleSet"  `
    -VirtualMachineScaleSet $vmss

If the upgrade policy on your scale set is manual, update your VM instances with Update-AzVmssInstance. This cmdlet applies the updated scale set configuration to the VM instances and installs your application.

Cloud-init is a widely used approach to customize a Linux VM as it boots for the first time. You can use cloud-init to install packages and write files, or to configure users and security. As cloud-init runs during the initial boot process, there are no extra steps or required agents to apply your configuration.

Cloud-init also works across distributions. For example, you don't use apt-get install or yum install to install a package. Instead you can define a list of packages to install. Cloud-init automatically uses the native package management tool for the distro you select.

For more information, including an example cloud-init.txt file, see Use cloud-init to customize Azure VMs.

To create a scale set and use a cloud-init file, add the --custom-data parameter to the az vmss create command and specify the name of a cloud-init file. The following example creates a scale set named myScaleSet in myResourceGroup and configures VM instances with a file named cloud-init.txt. Enter your own names as follows:

az vmss create \
  --resource-group myResourceGroup \
  --name myScaleSet \
  --image Ubuntu2204 \
  --orchestration-mode uniform \
  --upgrade-policy-mode automatic \
  --custom-data cloud-init.txt \
  --admin-username azureuser \
  --generate-ssh-keys

When new OS releases are available, you can use or build a new custom image and deploy OS upgrades to a scale set. Each VM instance is upgraded to the latest image that you specify. You can use a custom image with the application preinstalled, the Custom Script Extension, or PowerShell DSC to have your application automatically available as you perform the upgrade. You may need to plan for application maintenance as you perform this process to ensure that there are no version compatibility issues.

If you use a custom VM image with the application pre-installed, you could integrate the application updates with a deployment pipeline to build the new images and deploy OS upgrades across the scale set. This approach allows the pipeline to pick up the latest application builds, create and validate a VM image, then upgrade the VM instances in the scale set. To run a deployment pipeline that builds and deploys application updates across custom VM images, you could create a Packer image and deploy with Azure DevOps Services, or use another platform such as Spinnaker or Jenkins.

As you build and deploy applications to your scale sets, you can review the Scale Set Design Overview. For more information on how to manage your scale set, see Use PowerShell to manage your scale set.