Deployment Stacks: Azures Standard for Infrastructure as Code, Drift Detection, and Governance @MicrosoftReactor
Deployment Stacks: Azures Standard for Infrastructure as Code, Drift Detection, and Governance  @MicrosoftReactor
Uploaded November 2026 | Updated November 2026, 21 hours ago
Once an Azure environment is deployed, it rarely stays the way the code describes it: people change settings in the portal, delete resources by hand, and leave old resources running.

Deployment stacks are Azure's standard for managing infrastructure as code across that lifecycle, with drift detection and governance built in.

I am a product manager at Microsoft on Azure Bicep and owning Deployment Stacks. Bicep is Azure's native infrastructure as code language and a stateless alternative to Terraform, so there is no state file to store, secure, or keep in sync.

A deployment stack is a resource that manages a group of resources as a single unit across its lifecycle. In the demo I deploy an environment from Bicep as a deployment stack, then change a security rule and delete a resource in the portal. Stacks What-If with the same template lists those changes, and one more deployment puts the environment back. I then set deny settings so managed resources cannot be changed or deleted outside the stack, and choose whether resources removed from the template are detached or deleted.

When the environment is no longer needed, a single delete removes everything the stack manages. Everything runs live from the Azure CLI. The tools are Azure specific; the practices apply to any infrastructure as code workflow.

What attendees will learn:
- How a deployment stack manages an Azure environment as one unit, from the first deployment to the final delete.
- How to find drift with Stacks What-If and bring the environment back to what the code describes.
- How to govern changes with deny settings, and choose what happens to resources removed from the template.

[eventID:27816]
Deployment Stacks: Azures Standard for Infrastructure as Code, Drift Detection, and GovernanceReplacing Azure Communication Services Without Rebuilding Your ApplicationDe datos sin procesar a datos preparados para IA: Spark en FabricFrom GitHub to GPU: Build an Open Source AI App with Copilot, Azure Linux, and AKSAI @Work: Community Builders in ActionCopilot como tu primer asistente de IA: diseña un asistente para una tarea realModel Mondays - Spotlight On Fine-Tuned ModelsEpisode 2: Generative AI – Creating, Innovating, and Accelerating with AIEpisode 1: AI Fundamentals – Understanding the Future of WorkCapacidades avanzadas de MCP: Apps, Tasks y WebMCPSkill up on Microsoft Copilot StudioHow to trust the code you ship
Microsoft Reactor |

Deployment Stacks: Azure's Standard for Infrastructure as Code, Drift Detection, and Governance

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER