// This README explains The Azure VM image management project.

Purpose of the Project

Explore

It is an exploration of how a Virtual Machine can be deployed, imaged, and then recreated from that image. I want the recreated VM to have an ephemeral disk. So that I may measure the IOPS performance. I want azure to be used as the cloud service provider.

Use cases

• User runs pipeline 1 to create 1st generation VM from Azure Windows-latest image.
• User runs pipeline 2 to capture a generalized image of the 1st generation VM and to store it in a compute gallery.
• User runs pipeline 3 to create a 2nd generation VM from the compute gallery image.
• VM can be created from captured generalized image stored in Azure compute gallery.
• A second generation VM will persist until the user decides to approve its deletion.
• A user will need to be able to launch a powershell script remotely on the 2nd generation VM to be able to measure disk IOPS performance.

Technology Stack

Software development environment

Use bicep to code deployment of azure resources. Use powershell to code azure vm image capture to compute gallery. Use yml to code for az pipeline run.

Virtual Machine

Use OS Windows latest. 2 CPU of modest performance (enough to execute powershell scripts via terminal) No GPU required. 4 GB of RAM A first generation VM will use a managed disk. A second generation VM will use an ephemeral cache disk.

Network to VM for developer access

Allow for remote Az CLI powershell script run. No need to allow for remote desktop access.

Project Structure and Best Practices

1. Directory Structure

• infra/: Place all Bicep (IaC) files here for clarity and azd compatibility.
• scripts/: Place all PowerShell and automation scripts here.
• azure-pipelines.yml: CI/CD pipeline definition.

2. Security

• Never hardcode credentials. Use Azure Key Vault or pipeline secrets for sensitive values.
• Use managed identities for authentication where possible.

3. IaC (Bicep)

• Use parameter files for environment-specific values.
• Add comments to Bicep files for clarity.
• Use resource tags for traceability.

4. Automation

• Scripts should include error handling and logging.
• Use azd or az cli for deployments, but prefer azd for new projects.

5. Pipeline

• Store secrets in Azure DevOps variable groups or Key Vault.
• Use pipeline stages for infra, image, and VM deployment.
• It is expected for any pipeline that creates a VM resource, a cleanup stage will be run after approval by the user.

6. Azure resources

• Deploy to centralcanada location
• Use the cheapest VM SKU for the region.
• Encode resource deletion into pipeline cleanup stage activated after user approval.

7. Naming Conventions

• Use clear and descriptive names for resources.
• Use a consistent naming pattern, using env subscription, location and resource type. e.g., dcacevm1, dcacedisk, dcacecg.
• inspire yourself from the Azure naming conventions for resources.
• inspire yourself from Sumerian mythology for names of resources.

---