RLS is not working when I have multiple columns in the rules dataset

Q&A
,
1

I created a rules dataset and applied row-level security to my dataset using two columns: Partner and Client. When I assigned myself access to a specific Partner and Client, the Partner filter works correctly, but the Client filter does not — it still displays all clients.

My rules dataset has following columns which is exactly matching with my dataset:

GroupName,UserName,Partner Name,Client Name
Reporting,Nik****@abc.com,abacd,bcasb

2

Hi @Nikhil_1912

Welcome to the community.

Can you share the details of your dataset and the sample data where you say the rule does not apply?

Regards,
Giri

3

Thank you Giri,

Here is my dataset and the rules data set structure.

I’ve set up my dataset along with the corresponding rules dataset. While the row-level security works correctly for filtering by Partner Name, it doesn’t seem to apply to Client Name. Even though I’ve explicitly specified a value for Client Name in the rules dataset, the data still shows records for all clients.

Dataset:
Partner Name|Partner_id|Client Name|client_id|…
Abcdef|12345|GHIKL|5678|…

Rules Dataset:
GroupName|UserName|Partner Name|Client Name
Reporting|nik…@gmail.com|Abcdef|GHIKL

4

Hi @Nikhil_1912,

Is there another row in the rules dataset that’s possibly giving you access to all clients? For example, if you’re in the “Reporting” group and there’s a row like this, that would give you access to all clients, regardless of the other rules. When there are multiple rules for the same user, Quick Sight uses OR logic.

GroupName|UserName|Partner Name|Client Name
Reporting|null|null|null

5

Hello David,

Thanks for your response.

I checked that and this user is part of other group. Thanks for your help.