Why 3 tables are needed for RLS access setup?

[Anonymous]

I have a legacy SSAS cube where I see the below tables:

Tables description:

Table Access3: (The table is combination of project+user).

• Project Number
• User number
• CompositeKey (concatenation of Project+user)
• Flag*

Table Access2:

• CompositeKey

Table Access1:

• Role
• Project Number
• User number

Relationships:

Access3 <-> Access2 -> Access1 <-> Dimension -> Fact

• They connect through CompositeKey column.
• Connection Access1 <-> Dimension is through Project Number.

*This ‘flag’ is used in a .pbix that has an RLS and only displays specific projects for specific users dynamically.

My questions are:

• When the ‘flag’ column is not used, does it mean that that Access3 table/filter gets ignored? and the only applied filter is Access1?
• Are the 3 access tables needed to allow this dynamic behaviour where a cube can support RLS/personal reports and also less restrictive simultaneously?

[v-luwang-msft]

Hi  @Anonymous ,

Has your problem been solved, if so, please consider Accept a correct reply as the solution or share your own solution to help others find it.

Best Regards
Lucien

[amitchandak]

@Anonymous , Are you using userprincipalname() in this ? UserNumber = email Id ?

What is rule you are trying to configure