New report based on multiple semantic models of existing reports - Visualizations problems

[hugo_p_23]

I created a new report based on multiple semantic models of existing reports, all these reports have RLS.

When sharing this new report online, people are not able to see some visuals.

I tried to give Builder accesses to the old semantic models, I created identical accesses at the RLS level in the new report (I used 2 dimension tables to connect to all the semantic models, one for calendar and one for Cost Centers).

I can't understand why it doesn't work.

We are talking about 4 different semantic models, and about 50 different tables.

Thanks in advance.

Hugo

[AmiraBedh]

First thing that comes to my mind is an issue with RLS, so you need to verify if the new report match the ones in the individual semantic models because even slight differences in RLS logic could prevent users from accessing data.

Why ? because when you combine multiple semantic models, RLS might intersect or overlap, restricting some users unintentionally. For example, if a user has access in one model but not in another, visuals relying on restricted data might not render.

Even if users can access the new report, they also need sufficient permissions on the underlying semantic models (datasets). Builder access might not be enough they may also need specific permissions to view data.

I recommend that you test the RLS filters applied to your new report for each user group. 

Another thing, you have the model to model connections, since here you are using 2 dimension tables (Calendar and Cost Centers) to connect multiple semantic models and this can create ambiguity or break some relationships.