MS Fabric deploy changes on semantic model inplace

[marius1106]

Dear Support,

In MS Fabric I have 2 different Workspaces for a Dev- & Prod-Environment. In my Dev-Environment I edited my semantic Model and tested it. Now I want to deploy the changes to the Prod-Workspace, but in my Deployment Pipeline both semantic models are not linked so I can't update it (they are seen as to different objects althouth they have the same name). 

Is there a way to deploy my changes inplace so that the Prod-Reports build on that model don't break?

Thank you!

[v-lgarikapat]

Hi @marius1106 ,

I wanted to follow up and confirm whether you’ve had the opportunity to review the information  provided by @suparnababu8 @burakkaragoz .

We truly appreciate your continued engagement and thank you for being an active and valued member of the community.

If you're still experiencing any challenges, please don’t hesitate to reach out  we’d be more than happy to assist you further.

We look forward to hearing from you.

Best regards,
Lakshmi

[suparnababu8]

Hi @marius1106 

Yes, if you deply the chnages to PD. it will break the PD semantci model definately.  As a work around, you can delete the PD sematic model and then deploy from DV to PD., make sure if you don't impact the chnages in PD..then only go for this workaround.

Thank you!!

Did I answer your question? Mark my post as a solution!

Proud to be a Super User!

[marius1106]

Hi @suparnababu8 @burakkaragoz ,

thank you for your answers. I tried it but when I deleted the prod-model and deployed the dev-model to prod, all the reports build on the prod-model got deleted. How can I avoid that?

[v-lgarikapat · Accepted solution]

Hi @marius1106 ,

1.Use Overwrite Instead of Delete
Don’t delete the Prod model manually.

Instead, deploy the Dev model directly to Prod using the same name, and let Fabric overwrite the existing model.

This preserves report bindings  as long as the model structure (tables, columns, measures) remains compatible.

If the pipeline doesn’t recognize the models as linked, you’ll need to pair them manually first using PBIP or Git integration.

2. Backup Reports Before Any Model Deletion
Use the Power BI REST API or manually download .pbix files for all reports tied to the Prod model.

This gives you a rollback option if something goes wrong.

3. Use PBIP Format with Git Integration
Convert your semantic model to PBIP format.

Link both Dev and Prod workspaces to the same Git repo.

This ensures consistent lineage, allowing Fabric to treat the models as the same object across environments.

Solved: Re: Deployment Rules for Semantic Models - Change ... - Microsoft Fabric Community

Solved: Re: Deployment Rules for Semantic Models - Change ... - Microsoft Fabric Community

Manage Direct Lake semantic models - Microsoft Fabric | Microsoft Learn

Troubleshoot the Fabric lifecycle management tools. - Microsoft Fabric | Microsoft Learn

We truly appreciate your continued engagement and thank you for being an active and valued member of the community.

If you're still experiencing any challenges, please don’t hesitate to reach out  we’d be more than happy to assist you further.

We look forward to hearing from you.

Best regards,
Lakshmi

[v-lgarikapat]

Hi @marius1106 ,

I just wanted to check if your issue has been resolved. If you still have any questions or need help, feel free to reach out I’m happy to assist.

Thank you for being an active part of the community. Looking forward to hearing from you!

Best regards,
Lakshmi

[burakkaragoz]

Hi @marius1106 ,

Yeah, this is super annoying but happens all the time. Your semantic models aren't "paired" in the deployment pipeline, so it sees them as totally different things even though they have the same name.

What's going on: When you created the prod model separately instead of deploying from dev, the pipeline doesn't know they're supposed to be the same thing. So it won't overwrite - it'll just create duplicates.

Easiest fix (but scary): Delete the semantic model from prod workspace, then deploy from dev. I know it's terrifying to delete production stuff, but this actually works best because:

• The deployment creates a properly paired model
• All your reports automatically connect to the new model
• Everything stays linked properly

Less scary option: Keep your current prod model for now, deploy from dev anyway (you'll get a duplicate), then gradually move your reports over to the new properly-paired model. Once everything's moved, delete the old one.

Quick alternative: If you have the original .pbix file, just upload it again to prod workspace with the same name. This basically "resets" the connection.

The deployment pipeline is pretty good at keeping report connections intact, so your dashboards shouldn't break. But yeah, the pairing thing is one of those Fabric quirks that trips everyone up.

---

If my response resolved your query, kindly mark it as the Accepted Solution to assist others. Additionally, I would be grateful for a 'Kudos' if you found my response helpful.

This response was assisted by AI for translation and formatting purposes.