Integrate Power BI with Azure Devops

[AmiraBedh]

I have some questions regarding CI/CD in my case.

I am using a DEV workspace to develop the reports and semantic models.

The other workspaces TEST and PROD, I use deployment pipelines to deploy from DEV to TEST to PROD.

Question 1 : My thinking is I need only one repo for DEV and no need to have repos for TEST and PROD. Is this correct or considered a bad practice?

Question 2 : 

Let's say I have 3 workspaces for DEV, each workspace is for a specific subject (project) HR, Payroll and Timesheet

• HR DEV 
• Payroll DEV
• Timesheet DEV

and I use deployment pipelines to deploy the content for each DEV workspace into the TEST and PROD.

I think of having only one Devops project with (for the repository a main branch and 3 branches for each content coming from HR DEV, Payroll DEV and Timesheet DEV. Is it a good approach?

Or should I create a sperate branch for each workspace?

Or should I create a sperate project for each workspace ?

Question  3 :

IIs there any approval stategy in this case ?

Question 4 :

How is the release done in this case ?

[rajendraongole1 · Accepted solution]

Hi @AmiraBedh - 

Q1:

One repo for DEV is usually enough; TEST and PROD don’t need separate repositories unless special circumstances require it. (It's common to have a single repository for DEV and use deployment pipelines to promote content. You don't need separate repositories for TEST or PROD unless your organization requires strict version control at every stage.)

Q2:

Single DevOps project with branches for each workspace is a good strategy to manage content for HR, Payroll, Timesheet. (Once you finish developing a report or model in, say, the HR DEV branch, you can merge it into the main branch to promote to TEST and PROD via deployment pipelines.

This strategy helps maintain a clean structure and reduces complexity. Having separate DevOps projects for each workspace would add unnecessary overhead unless the projects are managed by entirely different teams.)

Q3:

Yes, you can implement an approval strategy within Azure DevOps, and this integrates well with Power BI deployment pipelines

Approval processes can be managed in Azure DevOps, with gates for manual approval before TEST and PROD. (Link shared below)

Releases follow the well-defined process (DEV → TEST → PROD) and can be controlled via DevOps pipelines with manual approval and testing.(You can automate this flow using Azure DevOps pipelines and deployment stages, but the manual approval step ensures quality control before promotion to higher environments.)

Reference links:

Overview of Fabric deployment pipelines - Microsoft Fabric | Microsoft Learn

Azure: Understand release gates, checks, and approvals - Azure Pipelines | Microsoft Learn

I hope the above details and reference links helps in implementing the same.