Forum Discussion
Segregate single Power BI Dataset for different workspaces
Greetings Power BI Community !
I am looking for best possible ways to maintain a single unified dataset across multiple workspaces.
For my case, I have a single Power BI Framework connected using DirectQuery mode. I have two separate workspaces; W1 and W2. W1 is designated for Product Category 'A' and the W2 for Product Category 'B' and 'C'.
The reports and dashboards are going to be the same. The ones published in W1 should only show the data belonging to Product Category 'A' and same set of reports should display the data belonging to Product Category 'B' and 'C' in W2 (This can be controlled through RLS)
But this is not limited to the reports and dashboards. Even for the contributors, the above set of rules should apply i:e W1 should only show Category 'A' data and W2 should only show data pertaining to Category 'B' and 'C'
We are looking to maintain a single unified Power BI Framework file. Apart from RLS which only applies to the viewer role, Is there a recommended approach to handle this case?
One way to control this is to have two separate views created at backend, where we explicitly filter out the Product Category as per the workspace requirement and expose that view in the PBI Framework for that specific workspace. This way, we end up with having two different framework files, which the business is trying to avoid.
Any recommendation on this will be highly appreciated. Please let me know if you require further information related to this case.
Thanks in Advance.
4 Replies
- TomMartensSuper User
Hey PBI_Member_01
you can achieve what you are looking for by using a 3rd workspace that only holds semantic model. Content creators who are publishing content reports (reports that hold the data visualizations) do not need to have a role assigned to this data workspace. The content creators do not connect to the underlying data source, instead the content creators are connecting to the Power BI dataset from inside Power BI Desktop.
The only thing content creators need is the build permission to the sematic model. This approach will give you:
- a single semantic model (a shared model)
- content creators can create their own reports with RLS honored
Hopefully, this provides what you are looking for to tackle your challenge.
Regards,
Tom
- PBI_Member_01Helper III
Hi TomMartens
Thank you for your suggestion.
I understand the approach you are taking but we only have two workspaces to work with and having a separate workspace for maintaining the datasets does not adhere with the organization's architecture. Something that is not in my control.
Currently, there is also RLS applied which gives access to users according to their specific roles and can only see their own hierarchy of data within the organization.
While keeping that intact, any other possible approach that you may suggest for this problem.
Best Regards- TomMartensSuper User
Hey PBI_Member_01 ,
not considering a separate workspace for the semantic model is wrong as sometimes this is essential. I never heard of a "there are only two workspaces available" architecture. From my perspective the organization errs.
I recommend reading
- the Power BI Guidance documentation: Power BI guidance documentation - Power BI | Microsoft Learn
Regards,
Tom