Forum Discussion

MaxSimes's avatar
MaxSimes
Regular Visitor
2 years ago

RLS on Upstream Semantic Model

Hi Hive Mind,

 

I have been wrestling with this for a few days and am not getting anywhere so figured it was tiem to ask.

 

I have  a semantic Model 'PEPI - NOHH' with RLS set up on the table 'Meterpoint Details'

 

this is working as expected and filters dwon the data on a report that is built from thos model.

 

 

I have now created a new report that is connected to this Model as an Upstream Semantic Model along with another Upstream Model.

 

 

 

in this set up the RLS on PEPI - NOHH is no longer working, i cant set up RLS in Desktop as the Report is only using existing Semantic Models....

 

how can i achieve the ame RLS filtration with this setup or is this not possible?

 

Thanks in advance

 

Max

1 Reply

  • Anonymous's avatar
    Anonymous
    Not applicable
    2 years ago

    Hi MaxSimes ,

     

    Hope everything is going well.

     

    In Power BI, RLS is designed to protect row-level data in a specific dataset or model.

     

    When a base semantic model is used as an upstream source in another report or model, the RLS rules defined in the base semantic model are not automatically applied. This is because the RLS context is specific to the dataset in which it is defined and is not propagated to reports or models that use the data indirectly.

     

    The alternative is to duplicate the RLS logic in the report or model you use.

     

    However, this approach has limitations, such as the inability to modify or apply RLS directly in Power BI Desktop due to the nature of the connection (for example, DirectQuery mode for semantic models).

     

    If you have any further questions please feel free to contact me.

     

    Best Regards,
    Yang
    Community Support Team

     

    If there is any post helps, then please consider Accept it as the solution  to help the other members find it more quickly.
    If I misunderstand your needs or you still have problems on it, please feel free to let us know. Thanks a lot!