Automating builds and deployments with Azure pipelines

You SHOULD use Azure Pipelines to automate build, test, and deployment processes, enabling Continuous Integration (CI) & Continuous Deployment (CD).

How do you organise your code?

Follow How to Organise Your Software Solution to maintain a clean, modular codebase and keep build and deployment scripts separate to reduce clutter.

Core pipeline characteristics

Follow Microsoft's baseline architecture to build repeatable, scalable Azure pipelines.

Key considerations:

• Use YAML: Prefer YAML pipelines for version control and sharing configurations.

• Modular Design: Create reusable templates & separate stages (e.g. Build, Test, Deploy).

• Gradual Deployment: Use progressive strategies (e.g. DevTest > UAT > Prod).

• Integrate Security: Include code analysis and vulnerability scanning.

• Automated Testing: Run automated tests with coverage reporting.

• Approvals

• Pre-approval checks are RECOMMENDED for DevTest and UAT.

• Pre-approval checks SHOULD be in place before deploying to Prod.

• Secure Secrets: Use tools like Azure Key Vault to manage secrets.

• Store Artifacts: Retain build outputs for traceability.

Prefer YAML over classic release pipelines

Use YAML pipelines for improved version control, tracking, and auditing.

While Classic Release pipelines can still be useful, YAML deployment jobs and Azure DevOps Environments are preferred for greater flexibility and scalability.

Deployment groups may suit VM-based deployments but offer less control than YAML.

Naming conventions

Link work items to builds

Automatically link work items included: In Pipeline settings choose to connect work items with each build result.