Forum Discussion
How to create a connection to a database with integrated credentials and then use it in PBI Desktop?
- 9 months ago
Hi Fern_21,
Users can ONLY access the tables you include in the Dataflow.
When users connect to your Dataflow in Power BI Desktop, they can:
Import only the entities (tables) inside the Dataflow
Build their own model on top of those tables
Combine with other local files (Excel, CSV, etc.) if they want
Hi Fern_21 ,
Directly creating a file (like a .pbix or .pbids) with embedded database credentials for another user to open and use in Power BI Desktop is not possible for security reasons.
To avoid giving access to database for all the users in organization, one person with database access can load
all the data in pbix in import mode and then share it with all users.in this approach, users will require access to semantic model not the database.this approach is known as self service bi.there are lot of articles on it.
https://learn.microsoft.com/en-us/power-bi/connect-data/service-datasets-manage-access-permissions
https://radacad.com/power-bi-user-access-levels-build-and-edit-are-different/
https://learn.microsoft.com/en-us/power-bi/connect-data/service-datasets-across-workspaces
Please give kudos or mark it as solution once confirmed.
Thanks and Regards,
Praful
Hi Praful_Potphode and grazitti_sapna !
If I create a semantic model and share it with the user, they can only connect in live connection mode and there are limitations. Isn't there a way to use a dataflow?
Thanks
- grazitti_sapna9 months ago
Super User
Fern_21, yes you can use a dataflow or fabric lakehouse to achieve it
- Fern_219 months ago
Helper V
Hi grazitti_sapna !
What is the fastest and most functional solution? What are the pros and cons of each?
Thanks- grazitti_sapna9 months ago
Super User
Hi Fern_21,
You can go for Dataflow,
Fast
Easy
Secure (no credentials exposed)
Cheapest
Lets users freely model their own semantic models
- Praful_Potphode9 months ago
Super User
Hi Fern_21 ,
Yes they can use dataflows to bring in user managed data and build composite models on top of it.
another apparoach can be to mirror the SQL Server data in fabric and then provide access inside fabric
https://learn.microsoft.com/en-us/fabric/mirroring/sql-server
please give kudos or mark it as solution once confirmed.
Thanks and Regards,
Praful