Check your eligibility for this 50% exam voucher offer and join us for free live learning sessions to get prepared for Exam DP-700.
Get StartedDon't miss out! 2025 Microsoft Fabric Community Conference, March 31 - April 2, Las Vegas, Nevada. Use code MSCUST for a $150 discount. Prices go up February 11th. Register now.
Dear community,
I have a question regarding the ownership of semantic models in a large organization. Currently, all our semantic models are owned by individual users, even those that are placed in "Production" workspaces.
Now, let's consider the following scenarios:
What is the recommended approach to avoid these kinds of situations?
Solved! Go to Solution.
Give permissions to the affected workspace(s) to user remaining in the company.
That users can "take over" the semantic models and afterwards (hopefully) renew the personal user-based cloud connections without issues. To renew personal use-based auth they need access to the sources of course.
Depends on how you have designed your access policies.
Regards.
This is a similar issue with large on-premises analysis service systems.
Do not have the ownership as a user.
Instead, create a 'service account' in Azure Active Directory (I think it is called Entra Id now) that an Power BI administrator has the password for. You need to assign a Power BI Pro license to this new user. Login to app.powerbi.com with the 'service account' credentials and Take Over by this user. Now, you need to have the 'service account' granted permissions to data sources in order to refresh the data.
If a normal user deployes a new model, it will overwrite the model but not the credentials.
Some poeple might say use a service principal, but service principals cannot be granted access to data sources, so you are still stuck with the same problem. Others might say to use credentials in a Data Gateway, still use a 'service account' for this.
This is a similar issue with large on-premises analysis service systems.
Do not have the ownership as a user.
Instead, create a 'service account' in Azure Active Directory (I think it is called Entra Id now) that an Power BI administrator has the password for. You need to assign a Power BI Pro license to this new user. Login to app.powerbi.com with the 'service account' credentials and Take Over by this user. Now, you need to have the 'service account' granted permissions to data sources in order to refresh the data.
If a normal user deployes a new model, it will overwrite the model but not the credentials.
Some poeple might say use a service principal, but service principals cannot be granted access to data sources, so you are still stuck with the same problem. Others might say to use credentials in a Data Gateway, still use a 'service account' for this.
What if you now have MFA, this limits the access of the service account to only one user, does anyone know of a solution for this? (or is this just job security)
The service account would have to have the MFA turned off. The IT Security or Infrastructure team would manage the credentials to insure no one uses the service account for anything else but the connection and semantic model ownership.
@3CloudThomas, @sergej_og, Thank you, guys! Both of your answers confirmed our thoughts. We now need to decide whether to create 'technical accounts' and determine how many of these accounts we need, considering the cost of the PRO license. Alternatively, we could continue using personal accounts as owners, but we would need to establish proper governance for individuals who leave the organization.
Give permissions to the affected workspace(s) to user remaining in the company.
That users can "take over" the semantic models and afterwards (hopefully) renew the personal user-based cloud connections without issues. To renew personal use-based auth they need access to the sources of course.
Depends on how you have designed your access policies.
Regards.
March 31 - April 2, 2025, in Las Vegas, Nevada. Use code MSCUST for a $150 discount!
Check out the January 2025 Power BI update to learn about new features in Reporting, Modeling, and Data Connectivity.
User | Count |
---|---|
19 | |
16 | |
15 | |
12 | |
11 |
User | Count |
---|---|
32 | |
25 | |
25 | |
19 | |
19 |