Forum Discussion

pmscorca's avatar
pmscorca
Icon for Post Prodigy rankPost Prodigy
9 months ago

Deployment order for warehouse object using deployment pipelines

Hi,

in order to deploy a warehouse having named schemas, tables and views using deployment pipelins I've noticed that it doesn't seem exist an order to perform a such task and this causes errors.

It is very important that a deployment pipeline runs following the dependency order between warehouse objects. In general:

  1. first, deploying schemas;
  2. second, deploying tables;
  3. third, deploying views.

I think that deployment pipelines can be a very useful feature but they need to improve without having to use DevOps or Git.

Thanks

 

CICD

3 Replies

  • Zanqueta's avatar
    Zanqueta
    Icon for Super User rankSuper User
    9 months ago

    Hi pmscorca,

     

    You are absolutely correct: deployment pipelines in Fabric currently do not enforce dependency order for warehouse objects (schemas → tables → views). This is a known limitation because pipelines treat objects as independent artefacts, and the deployment engine does not analyse dependencies.
     

    ✅ Recommended Workarounds

    1. Manual Sequencing in Pipelines
      • Break the deployment into three steps:
        • Step 1: Deploy schemas.
        • Step 2: Deploy tables.
        • Step 3: Deploy views.
      • This can be done by creating separate pipeline stages or using deployment rules to exclude certain objects in each stage.
    2. Use SQL Scripts with Ordered Execution
      • Instead of relying on object-level deployment, use a single SQL script that:
        • Creates schemas first.
        • Creates tables next.
        • Creates views last.
      • Add this script as a pipeline task in the deployment stage.
    3. Alternative: Git Integration
      • If you use Git integration, you can manage dependencies in your repository and control the order via CI/CD tools (Azure DevOps or GitHub Actions).
      • This is currently the most robust option for complex dependency management.

     

    Please, see the documentation:

    Overview of Fabric deployment pipelines - Microsoft Fabric | Microsoft Learn

     

     

    If this response was helpful in any way, I’d gladly accept a 👍much like the joy of seeing a DAX measure work first time without needing another FILTER.

    Please mark it as the correct solution. It helps other community members find their way faster (and saves them from another endless loop 🌀.

  • Anonymous's avatar
    Anonymous
    Not applicable
    8 months ago

    Hi pmscorca ,

     

    Thanks for creating this thread and sharing your thoughts on the deployment pipeline improvements. We really appreciate you bringing this to the community. If you have any more questions or need help with anything related to this topic, please feel free to reach out to us.

     

    Thank you.