Auto scaling App Service in Azure - Mini Project: Build & Apply

Auto scaling App Service

📖 Scenario: You are managing a web application hosted on Azure App Service. To handle varying user traffic efficiently, you want to set up auto scaling. This means the app will automatically add or remove instances based on CPU usage, so it stays fast and cost-effective.

🎯 Goal: Build an Azure Resource Manager (ARM) template snippet that configures auto scaling for an Azure App Service plan. The scaling rule should increase instances when CPU usage is above 70% and decrease when below 30%.

📋 What You'll Learn

Create a resource group variable named resourceGroupName with the value MyResourceGroup.

Create a variable named appServicePlanName with the value MyAppServicePlan.

Define an autoscale setting resource named autoscaleSetting targeting the App Service plan.

Configure two scale rules: one to scale out when CPU > 70%, another to scale in when CPU < 30%.

💡 Why This Matters

🌍 Real World

Auto scaling helps web apps handle traffic spikes smoothly without manual intervention, saving cost and improving user experience.

💼 Career

Cloud engineers and DevOps professionals often configure autoscaling to ensure applications remain responsive and cost-efficient.

1

DATA SETUP: Define resource group and App Service plan variables

Create two variables: resourceGroupName set to "MyResourceGroup" and appServicePlanName set to "MyAppServicePlan".

Azure

# Your code here

Need a hint?

Use var to declare variables and assign the exact string values.

2

CONFIGURATION: Define autoscale setting resource basics

Create a JSON object named autoscaleSetting with name set to "autoscaleMyAppServicePlan", type set to "Microsoft.Insights/autoscaleSettings", and targetResourceUri set to the resource ID of the App Service plan using resourceGroupName and appServicePlanName.

Azure

var resourceGroupName = "MyResourceGroup";
var appServicePlanName = "MyAppServicePlan";

// Your code here

Need a hint?

Use template literals with ${variable} to build the targetResourceUri.

3

CORE LOGIC: Add scale rules for CPU thresholds

Inside autoscaleSetting, add a profiles array with one profile named "AutoScaleProfile". Add two rules: one to increase capacity by 1 when CPU percentage is greater than 70, and one to decrease capacity by 1 when CPU percentage is less than 30. Use metric triggers on Percentage CPU and set timeAggregation to Average.

Azure

var resourceGroupName = "MyResourceGroup";
var appServicePlanName = "MyAppServicePlan";

var autoscaleSetting = {
  name: "autoscaleMyAppServicePlan",
  type: "Microsoft.Insights/autoscaleSettings",
  targetResourceUri: `/subscriptions/{subscriptionId}/resourceGroups/${resourceGroupName}/providers/Microsoft.Web/serverfarms/${appServicePlanName}`
  // Your code here
};

Need a hint?

Follow the Azure autoscale JSON schema for rules. Use metricTrigger and scaleAction objects inside each rule.

4

COMPLETION: Add location and finalize autoscale setting resource

Add a location property to autoscaleSetting with the value "eastus". Also add enabled set to true and notifications as an empty array. This completes the autoscale setting resource configuration.

Azure

var resourceGroupName = "MyResourceGroup";
var appServicePlanName = "MyAppServicePlan";

var autoscaleSetting = {
  name: "autoscaleMyAppServicePlan",
  type: "Microsoft.Insights/autoscaleSettings",
  targetResourceUri: `/subscriptions/{subscriptionId}/resourceGroups/${resourceGroupName}/providers/Microsoft.Web/serverfarms/${appServicePlanName}`,
  profiles: [
    {
      name: "AutoScaleProfile",
      capacity: {
        minimum: "1",
        maximum: "5",
        default: "1"
      },
      rules: [
        {
          metricTrigger: {
            metricName: "Percentage CPU",
            metricNamespace: "",
            metricResourceUri: `/subscriptions/{subscriptionId}/resourceGroups/${resourceGroupName}/providers/Microsoft.Web/serverfarms/${appServicePlanName}`,
            timeGrain: "PT1M",
            statistic: "Average",
            timeWindow: "PT5M",
            timeAggregation: "Average",
            operator: "GreaterThan",
            threshold: 70
          },
          scaleAction: {
            direction: "Increase",
            type: "ChangeCount",
            value: "1",
            cooldown: "PT5M"
          }
        },
        {
          metricTrigger: {
            metricName: "Percentage CPU",
            metricNamespace: "",
            metricResourceUri: `/subscriptions/{subscriptionId}/resourceGroups/${resourceGroupName}/providers/Microsoft.Web/serverfarms/${appServicePlanName}`,
            timeGrain: "PT1M",
            statistic: "Average",
            timeWindow: "PT5M",
            timeAggregation: "Average",
            operator: "LessThan",
            threshold: 30
          },
          scaleAction: {
            direction: "Decrease",
            type: "ChangeCount",
            value: "1",
            cooldown: "PT5M"
          }
        }
      ]
    }
  ]
  // Your code here
};

Need a hint?

Add location, enabled, and notifications properties at the top level of autoscaleSetting.