Strange issue with DAX formula and Row Level Security Role

[mstone3]

Hello -

I have a basic table below.  The DAX formula for the measure is a simple COUNTROWS function:

Measure =COUNTROWS(Losses)

I have a couple of blanks in the table and need those to be replaced with zeroes, so I add +0 to the formula and it works as desired:

Measure = COUNTROWS(Losses)+0

However, when I apply a row level security role and view the visual as that role, the undesired blanks return, and the DAX formula has not been changed, it is still Measure = COUNTROWS(Losses)+0

This is very strange - I don't understand what's going on here.  Thanks very much for your help!

[Greg_Deckler]

@mstone3 What is your RLS role definition in terms of its DAX formula?

[mstone3]

Hi @Greg_Deckler , 

The DAX formula for the role definition is [Level 2] = "NCTR"

Every row in the data has a 4 letter organizational acronym with column name "Level 2", so the role definition is just equal to the acronym. 

Thank you

[Greg_Deckler]

@mstone3 So would all of those rows for FY17 Involuntary Losses be filtered out for NCTR role? I'll have to do some mocking up of data unless you can provide a sample in text.

[mstone3]

@Greg_Deckler Yes that's correct - all of those rows would be filtered out by the NCTR role.  Based on your question, it sounds like there is a difference between the way RLS works and the way a regular slicer/filter would work without a RLS role applied, such that:

+With an RLS role enabled for 'NCTR', there are no rows/no data to access, so nothing to compute with the COUNTROWS function and therefore no +0 to add, so therefore blank results.

+If no RLS role, but instead filtered to 'NCTR' using a slicer/filter, there are still rows available for the formula to 'see' and for the COUNTROWS function to compute and then add +0, so therefore result shows zero. (all of the tables shown in the thread were filtered to 'NCTR' using a slicer)

Does that sound like the right way to think about it?

Thanks!