Skip to main content
  1. Public practice questions
  2. AZ-104 Practice Exams | Azure Administrator Associate
  3. AZ-104 Case Study Set

Question 42 — AZ-104 Case Study Set

Paideffort Enterprise Virtual Machine Operations Modernization
 

Business Context

Paideffort operates a global certification-learning platform hosted in Azure. The production environment contains application servers, batch-processing workers, administrative jump hosts, and specialized virtual machines that support content rendering and exam-processing workloads. The company currently manages 18 virtual machines but expects this number to exceed 70 during the next expansion phase.

Development teams also operate an AI-assisted certification service. Microsoft Foundry project endpoints are used for agent and project operations, while direct model requests use an Azure OpenAI endpoint. Microsoft Entra ID is required for supported authentication flows. FunctionTool handles deterministic operations, an OpenAPI tool retrieves certification-catalog data, and Azure Functions execute long-running document-processing tasks.

Operational requirements

Paideffort must be able to reproduce staging and production VM environments consistently. VM names, network names, sizes, and environment-specific settings must vary without requiring separate infrastructure definitions. Engineers need interactive command-line tools for occasional administrative work, repeatable automation for routine operations, and programmatic VM provisioning from an internal management application that contains branching business logic. Selected post-deployment configuration must run automatically on individual VMs.

Infrastructure Deployment

Paideffort has exported an existing application VM configuration and converted it into a reusable ARM template. The platform team plans to use the same infrastructure definition for staging and production by supplying different parameter values. Resource Manager is expected to reconcile deployed resources with the template when the definition is updated and redeployed.

{
  "$schema": "https://schema.management.azure.com/schemas/2019-04-01/deploymentTemplate.json#",
  "contentVersion": "1.0.0.0",
  "parameters": {
    "vmName": {
      "type": "string"
    },
    "vmSize": {
      "type": "string",
      "defaultValue": "Standard_D2s_v5"
    },
    "virtualNetworkName": {
      "type": "string"
    }
  },
  "resources": [
    {
      "type": "Microsoft.Compute/virtualMachines",
      "apiVersion": "2024-07-01",
      "name": "[parameters('vmName')]",
      "location": "[resourceGroup().location]",
      "properties": {
        "hardwareProfile": {
          "vmSize": "Standard_D4s_v5"
        },
        "networkProfile": {
          "networkInterfaces": [
            {
              "id": "[resourceId('Microsoft.Network/networkInterfaces', parameters('nicName'))]"
            }
          ]
        }
      }
    }
  ]
}

The staging environment is expected to use Standard_D2s_v5, while production administrators may choose a larger supported VM size. The networking team also wants the network interface name supplied independently for each environment.

Administrative Operations

Cloud engineers frequently create temporary Linux VMs for troubleshooting. Most administrators work from Windows or Linux workstations, and some connect through Azure Cloud Shell when a local command-line environment is unavailable. One team uses the following provisioning step for short-lived Linux workers:

az vm create \
  --resource-group Paideffort-Operations \
  --name pdft-temp-worker-01 \
  --image Ubuntu2204 \
  --admin-username azureuser \
  --generate-ssh-keys

Another administrator maintains recurring VM maintenance scripts in PowerShell. These scripts use Azure PowerShell cmdlets and are run interactively during investigations or scheduled when the same operation must be repeated across a controlled group of machines.

Application-Driven Provisioning

Paideffort is also developing an internal capacity-management application. The application evaluates course-release schedules, queue depth, regional capacity, and licensing requirements before determining whether additional VMs should be provisioned. The developers prefer a higher-level programming abstraction instead of constructing and processing raw HTTP requests for every Compute operation.

var vmName = requestedVmName;

azure.VirtualMachines.Define(vmName)
    .WithRegion(Region.USEast)
    .WithExistingResourceGroup("Paideffort-Production")
    .WithExistingPrimaryNetworkInterface(networkInterface)
    .WithLatestWindowsImage(
        "MicrosoftWindowsServer",
        "WindowsServer",
        "2012-R2-Datacenter")
    .WithAdminUsername("platformadmin")
    .WithAdminPassword("Paideffort-Prod-Password")
    .WithComputerName(vmName)
    .WithSize(VirtualMachineSizeTypes.StandardDS1)
    .Create();

Post-Deployment Configuration

Several worker VMs require a monitoring agent and a Paideffort processing package after the operating system has been provisioned. The configuration must execute automatically after deployment and should be associated directly with the target VM rather than requiring administrators to sign in interactively.

Current operations workflow

An administrator creates each VM, signs in through SSH or RDP, downloads the required software, installs the monitoring package, applies application settings, verifies configuration, and repeats the procedure on the next VM. The operations team expects approximately 40 new workers during the next deployment window.

Automation and Lifecycle Management

Paideffort's operations center monitors VMs for a recurring application-service failure. When a specific event is detected, the organization wants an automated workflow to respond without waiting for an administrator. The infrastructure team also needs centralized visibility into operating-system updates, the ability to schedule patch installation, and deployment results that show whether updates were applied successfully.

- environment: development
  vmCount: 12
  shutdown:
    enabled: true
    schedule: "19:00"
    timezone: "GMT Standard Time"

- environment: production
  vmCount: 28
  shutdown:
    enabled: true
    schedule: "19:00"
    timezone: "GMT Standard Time"

Development workers are normally unused overnight, so the finance team expects them to shut down automatically outside working hours. Production workers, however, process certification workloads continuously and are expected to remain available overnight.

Future requirements

Paideffort expects its VM fleet to grow across additional Azure regions. Infrastructure teams want standardized repeatable provisioning, application developers want programmatic control through supported SDKs, and administrators need automated post-deployment configuration. Operations leadership wants fewer manual maintenance procedures, lower nonproduction compute cost, scheduled patch management, event-driven remediation, stronger credential practices, and centralized operational reporting.