Changing server/database for multiple semantic models without republishing

[Mohamadmonem]

Hi Everyone!

We have multiple semantic models in Power BI Service connected through an On-premises Data Gateway.

If the database server address or database name changes, is there a way to update the connection centrally without opening and republishing each semantic model?

Can this be handled by:

• Updating the Gateway data source only?

• Using parameters?

• Using the REST API (Update Datasource)?

• Or is updating each semantic model individually unavoidable?

We want to avoid updating each semantic model one by one.

[MohdZaid_ · Accepted solution]

Hey @Mohamadmonem  , 

You can use the Update Datasources REST API to update multiple semantic models programmatically:

https://learn.microsoft.com/en-us/rest/api/power-bi/datasets/update-datasources

This is ideal if:

You have many models

You want to avoid manual work

You’re comfortable using PowerShell or automation

If it solved your issue, feel free to mark it as the solution so others can benefit too.

Thanks for being part of the community.

[cengizhanarslan]

You can define parameters for your Server and Database info in PowerQuery and use those parameters as a part of table source. However you would still need to individually change parameters in semantic model options in power bi service.

[Mohamadmonem]

Thanks for your reply.

I understand that parameters can be used if they are defined in Power Query from the beginning.

However, in our case, the semantic models were not originally built with parameters. That means we would need to modify and republish each model to introduce them.

Is there any alternative way to update the server/database connection centrally without rebuilding and republishing each semantic model?