Get certified for free when you join Fabric Data Days 2026 and dive into Fabric, Power BI, SQL, AI, and other essential data skills.
Join nowTry your skills in the Power BI Dataviz World Championship! Round one ends June 26. Join now
Hello, we are trying to use a single dimension table with RLS to filter multiple fact tables. It worked previously, and we have a model that it still works in. However, everything looks the same from a relationship perspective. Many-to-many bidirectional security filter is not enabled. Here is a video explaining what we are seeing.
@Anonymous FYI
Sorry, I never came back to this. We found an underlying issue with inbound data and relationships.
And, we are definitely aware that the model is far from ideal. Inherited it. Likely going to completely re-write this in the very near future.
@crisher , I see three tables related to many to many bi-directional joins, Not an ideal model.
Many to many fine if those two tables using RLS using user.
Can you share a sample pbix after removing sensitive data.
refer if needed
How to use Row Level Security (RLS): https://youtu.be/NfdIA0uS6Nk
| User | Count |
|---|---|
| 23 | |
| 21 | |
| 20 | |
| 17 | |
| 11 |
| User | Count |
|---|---|
| 55 | |
| 53 | |
| 41 | |
| 36 | |
| 32 |