Forum Discussion
RLS Not Working with "Intermediary" Model
Apologies if this is confusing...
I have a "core" semantic model called "Financial Model". It has RLS implemented within it. It works fine when tested within Power BI Desktop and also works with a very basic "RLS Test" report on the Power BI service. However, I have another report called "Expense Report" that is also built on top of the "Financial Model" model and RLS does not work within it.
My theory is that because "Expense Report" has some custom measures, when published, a new "intermediary" model called "Expense Report" is published alongside it and RLS is not being passed from the "Financial Model" model, up through to the "Expense Report" model, and ultimately to the "Expense Report" report. Hopefully the image below helps.
Can this be fixed? In other words, can I have a report with custom measures, built on a model with RLS, and still have RLS passed through to my report?
I got to the bottom of this. At some point, one of the developers enabled DirectQuery, which creates this "intermediary" semantic model. I don't think there's a way to remove that so we are copying and pasting everything from the original report to the new report.
10 Replies
There isn't a semantic model explicitly/intentionally being created on top of it. I'm only guessing that it's being created because the report has some custom columns or measures in it. The intention is for the report to sit directly on top of the semantic model with RLS. Is there a way to resolve this?
- aj1973
Community Champion
You can build a new report with live connection to your dataset
I got to the bottom of this. At some point, one of the developers enabled DirectQuery, which creates this "intermediary" semantic model. I don't think there's a way to remove that so we are copying and pasting everything from the original report to the new report.
