Environment and Workload Isolation with CI/CD

Hi Anonymous ,

Thanks for reaching out to the Microsoft fabric community forum. 

 

Microsoft Fabric's deployment pipelines are designed to facilitate the promotion of content across different environments—such as development (DEV), testing (TEST), and production (PROD). However, as of now, deployment pipelines are confined to workspaces within a single capacity. This means that while you can manage deployments across different stages (DEV, TEST, PROD), all these stages must reside within the same Fabric capacity.​

This design choice stems from the current architecture of deployment pipelines in Microsoft Fabric, which operate independently of the capacities assigned to your workspaces. When a pipeline runs, orchestration activities (e.g., triggering deployments) are billed to the source workspace’s capacity. After deployment, activities in the target workspace utilize the target workspace’s capacity. The pipeline itself does not consume resources at this stage. ​

Regarding the use of multiple capacities within deployment pipelines, each workspace assigned to a pipeline stage must reside on a Fabric capacity (or use a Premium Per User license). You can use any supported capacity tier, including smaller F capacities, as long as the workspace is on a Fabric (Premium) capacity. ​

 

In scenarios where isolating environments across different capacities is a best practice—such as separating development, testing, and production workloads to ensure optimal performance and resource allocation—you might consider the following approaches:​

Manual Deployment Management: Manually export content from one capacity and import it into another. This process can be automated using scripts or external tools, but it requires additional management overhead.​

Capacity Management Pipelines: Utilize Fabric's Data Factory pipelines to orchestrate capacity operations, such as pausing, resuming, or scaling capacities. This approach allows for some level of automated capacity management, though it may not fully address the need for deployment pipelines across multiple capacities. ​

It's important to note that these methods may not provide the seamless deployment experience that native deployment pipelines within a single capacity offer. Therefore, when designing your deployment strategy, consider the trade-offs between capacity isolation benefits and the operational complexity introduced by managing deployments across multiple capacities.

Can you please refer Microsoft's official documentation and community forums.

https://learn.microsoft.com/en-us/azure/architecture/analytics/architecture/fabric-deployment-patterns
https://learn.microsoft.com/en-us/fabric/data-factory/cicd-pipelines
https://learn.microsoft.com/en-us/fabric/cicd/deployment-pipelines/intro-to-deployment-pipelines?tabs=new-ui
https://learn.microsoft.com/en-us/fabric/cicd/best-practices-cicd
https://powerbi.microsoft.com/en-in/blog/deployment-pipelines-azure-devops-extension-multiple-pipelines-working-together-dataflows-ga/
https://learn.microsoft.com/en-us/power-platform-release-plan/2021wave2/power-bi/deployment-pipelines-multiple-pipelines-working-together
Solved: Microsoft Fabric deployment pipelines - Microsoft Fabric Community

Solved: Fabric Capacity for Deployment Pipeline - Microsoft Fabric Community

 

If you find this post helpful, please mark it as an "Accept as Solution" and consider giving a KUDOS. Feel free to reach out if you need further assistance.
Thanks and Regards