Quickstart: Create a Linux virtual machine scale set with an Azure template

A virtual machine scale set allows you to deploy and manage a set of identical, auto-scaling virtual machines. You can scale the number of VMs in the scale set manually, or define rules to autoscale based on resource usage like CPU, memory demand, or network traffic. An Azure load balancer then distributes traffic to the VM instances in the scale set. In this quickstart, you create a virtual machine scale set and deploy a sample application with an Azure Resource Manager template.

If you don't have an Azure subscription, create a free account before you begin.

Open Azure Cloud Shell

Azure Cloud Shell is a free, interactive shell that you can use to run the steps in this article. Common Azure tools are preinstalled and configured in Cloud Shell for you to use with your account. Just select the Copy button to copy the code, paste it in Cloud Shell, and then press Enter to run it. There are a few ways to open Cloud Shell:

Select Try It in the upper-right corner of a code block. Cloud Shell in this article
Open Cloud Shell in your browser. https://shell.azure.com/bash
Select the Cloud Shell button on the menu in the upper-right corner of the Azure portal. Cloud Shell in the portal

If you choose to install and use the CLI locally, this tutorial requires that you are running the Azure CLI version 2.0.29 or later. Run az --version to find the version. If you need to install or upgrade, see Install Azure CLI 2.0.

Define a scale set in a template

Azure Resource Manager templates let you deploy groups of related resources. Templates are written in JavaScript Object Notation (JSON) and define the entire Azure infrastructure environment for your application. In a single template, you can create the virtual machine scale set, install applications, and configure autoscale rules. With the use of variables and parameters, this template can be reused to update existing, or create additional, scale sets. You can deploy templates through the Azure portal, Azure CLI 2.0, or Azure PowerShell, or from continuous integration / continuous delivery (CI/CD) pipelines.

For more information on templates, see Azure Resource Manager overview

To create a scale with a template, you define the appropriate resources. The core parts of the virtual machine scale set resource type are:

Property Description of property Example template value
type Azure resource type to create Microsoft.Compute/virtualMachineScaleSets
name The scale set name myScaleSet
location The location to create the scale set East US
sku.name The VM size for each scale set instance Standard_A1
sku.capacity The number of VM instances to initially create 2
upgradePolicy.mode VM instance upgrade mode when changes occur Automatic
imageReference The platform or custom image to use for the VM instances Canonical Ubuntu Server 16.04-LTS
osProfile.computerNamePrefix The name prefix for each VM instance myvmss
osProfile.adminUsername The username for each VM instance azureuser
osProfile.adminPassword The password for each VM instance P@ssw0rd!

The following example shows the core scale set resource definition. To customize a scale set template, you can change the VM size or initial capacity, or use a different platform or a custom image.

  "type": "Microsoft.Compute/virtualMachineScaleSets",
  "name": "myScaleSet",
  "location": "East US",
  "apiVersion": "2017-12-01",
  "sku": {
    "name": "Standard_A1",
    "capacity": "2"
  "properties": {
    "upgradePolicy": {
      "mode": "Automatic"
    "virtualMachineProfile": {
      "storageProfile": {
        "osDisk": {
          "caching": "ReadWrite",
          "createOption": "FromImage"
        "imageReference":  {
          "publisher": "Canonical",
          "offer": "UbuntuServer",
          "sku": "16.04-LTS",
          "version": "latest"
      "osProfile": {
        "computerNamePrefix": "myvmss",
        "adminUsername": "azureuser",
        "adminPassword": "P@ssw0rd!"

To keep the sample short, the virtual network interface card (NIC) configuration is not shown. Additional components, such as a load balancer, are also not shown. A complete scale set template is shown at the end of this article.

Add a sample application

To test your scale set, install a basic web application. When you deploy a scale set, VM extensions can provide post-deployment configuration and automation tasks, such as installing an app. Scripts can be downloaded from Azure storage or GitHub, or provided to the Azure portal at extension run-time. To apply an extension to your scale set, you add the extensionProfile section to the preceding resource example. The extension profile typically defines the following properties:

  • Extension type
  • Extension publisher
  • Extension version
  • Location of configuration or install scripts
  • Commands to execute on the VM instances

The Python HTTP server on Linux template uses the Custom Script Extension to install Bottle, a Python web framework, and a simple HTTP server.

Two scripts are defined in fileUris - installserver.sh, and workserver.py. These files are downloaded from GitHub, then commandToExecute runs bash installserver.sh to install and configure the app:

"extensionProfile": {
  "extensions": [
      "name": "AppInstall",
      "properties": {
        "publisher": "Microsoft.Azure.Extensions",
        "type": "CustomScript",
        "typeHandlerVersion": "2.0",
        "autoUpgradeMinorVersion": true,
        "settings": {
          "fileUris": [
          "commandToExecute": "bash installserver.sh"

Deploy the template

You can deploy the Python HTTP server on Linux template with the following Deploy to Azure button. This button opens the Azure portal, loads the complete template, and prompts for a few parameters such as a scale set name, instance count, and admin credentials.

Deploy template to Azure

You can also use the Azure CLI 2.0 to install the Python HTTP server on Linux with az group deployment create as follows:

# Create a resource group
az group create --name myResourceGroup --location EastUS

# Deploy template into resource group
az group deployment create \
    --resource-group myResourceGroup \
    --template-uri https://raw.githubusercontent.com/Azure/azure-quickstart-templates/master/201-vmss-bottle-autoscale/azuredeploy.json

Answer the prompts to provide a scale set name, instance count, and admin credentials for the VM instances. It takes a few minutes for the scale set and supporting resources to be created.

Test your scale set

To see your scale set in action, access the sample web application in a web browser. Obtain the public IP address of the load balancer with az network public-ip list as follows:

az network public-ip list \
    --resource-group myResourceGroup \
    --query [*].ipAddress -o tsv

Enter the public IP address of the load balancer in to a web browser in the format http://publicIpAddress:9000/do_work. The load balancer distributes traffic to one of your VM instances, as shown in the following example:

Default web page in NGINX

Clean up resources

When no longer needed, you can use az group delete to remove the resource group, scale set, and all related resources as follows. The --no-wait parameter returns control to the prompt without waiting for the operation to complete. The --yes parameter confirms that you wish to delete the resources without an additional prompt to do so.

az group delete --name myResourceGroup --yes --no-wait

Next steps

In this quickstart, you created a Linux scale set with an Azure template and used the Custom Script Extension to install a basic Python web server on the VM instances. To learn more, continue to the tutorial for how to create and manage Azure virtual machine scale sets.