How to update semantic model's cloud connection after updating the Fabric workspace with Git API

[dsuneetha]

I have our Fabric dev workspace integrated with github repo Dev branch.  QA workspace is integrated with the same github Repo's Main branch.

I created a lakehouse, couple of delta parquet tables and a semantic model and  committed these changes from Dev Fabric workspace to git hub repo Dev branch and merged theses changes into Main branch using a PR.

Later I updated QA workspace with commits pushed to the Main branch using Fabric Git API. 

In QA workspace, the semantic model  is connecting to Dev workspace. I could not figure out how to update this connection. 

How to update the Semantic model to point to QA workspace's SQL anlytics end point using the UI? Also is there any API to update this programmatically?

[dsuneetha]

Thank you for suggesting a solution. I also spoke with the support team. Currently there is no API to update just the SQL Analytics endpoint/lakeshouse id. Either we have to manually update model.ibm file and commit the change to git repo or use getItem/update Item definitions API to update it.

[v-shamiliv]

Hi @dsuneetha 
Thank you for your patience. From the steps outlined above, please let us know if anything was helpful to you, so that we can convert it into a formal answer. If so, we would appreciate it if you could Accept it as a solution and drop a 'Kudos' so other members can find it more easily.
Thank you.

[v-shamiliv · Accepted solution]

Hi @dsuneetha 

Thanks for reaching out Microsoft Fabric community forum

Currently, there is no direct UI option in Power BI or Fabric to update a semantic model’s connection to a new SQL Analytics endpoint directly in the Fabric workspace.

If the semantic model is linked automatically to a Lakehouse via Direct Lake mode and created natively in Fabric, there isn't a straightforward UI-based edit for its connection string. The workspace connection defaults to the workspace where the Lakehouse resides.

To move the semantic model:

Create a new semantic model in the QA workspace.

Use the same tables, measures, and structures as in the Dev model.

Connect the new model to the QA Lakehouse.

You can also update the connection programmatically by interacting with the Git repository where the configuration files for your semantic model are stored

If this post helps, then please give us Kudos and consider Accept it as a solution to help the other members find it more quickly.

Thank you.