Microsoft Fabric Community Conference 2025, March 31 - April 2, Las Vegas, Nevada. Use code FABINSIDER for a $400 discount.
Register nowGet inspired! Check out the entries from the Power BI DataViz World Championships preliminary rounds and give kudos to your favorites. View the vizzies.
Currently have RLS working excellently! Trying to reduce the dataset size for it as every report that external people see, uses it. I want RLS to work as is for external users but want to modify the security table so I don't have to have every single combination of internal employees and SecurityGroups. We have a lot! So, I added a column to the SecurityTable to flag if they are our employee or not. The table includes all external people and every security group they have. It also include a single record for every one of our employees with 1 for InternalEmployee.
TestTable
FirstName | SecurityGroup |
Bob | Group1 |
Tom | Group2 |
SecurityTable
Username | SecurityGroup | InternalEmployee |
susan@outside.org | Group1 | 0 |
tiffany@mycompany.com | NULL | 1 |
The relationship between these two tables is preventing me from doing this as the internal employee doesn't have a securitygroup. I have SecurityTable filtering TestTable joining on SecurityGroup. Is there a way to make RLS work this way listing tiffany just once or do I need to add a record in the security table for Group1, Group2, Group3 and so on?
For this example, Tiffany would see all the FirstNames and Susan would only see Bob.
Do you want to keep these tables connected or are you planning to use the RLS table in lookup mode? When you define the DAX rules for the RLS roles you can use your OR conditional as needed. Might also get away with COALESCE.
User | Count |
---|---|
86 | |
82 | |
42 | |
40 | |
35 |