Using RLS within Power BI to filter users

[Anonymous]

We have a report which is driven from CSV files & SharePoint lists (no SQL, no domain even!).

The report shows outstandings for each of our Centres around the country.

Up until recently I have had to have a report for each Centre (as other Centres shouldn't see each others data) which is a maintenance nightmare.....

I would like to be able to drive the report using Username() to filter by Centre with a table with email address and Centre in it.

I've looked at RLS and if I create a Role then the email address of the user is returned. However (and I believe this is by design at present as it is in Beta) when I refresh the dataset the Role is deleted.

That's no big deal (as long as a simple recreate of the Role would work) as updates to the reports are few and far between, but in my testing I am getting this error when I update the dataset and try to recreate the Role (called simply Staff with no members):

Trying to create a Role with a different name doesn't seem to work either.

Anybody got any ideas as to whether what I am trying to do will work or whether there is a better way to accomplish it?

Thanks.

Ian

[ankitpatira]

@Anonymous as you've said you are creating role with no members in it, try creating a role with members and filter in it. I believe it should work then as you would have added members and supplied DAX to filter rows. Check out this blog post on implementing RLS.

[Anonymous]

Thanks.

If I create the roles with users, will they still get deleted when the report is uploaded again?

If so, this isn't a great deal of use in a production environment. 😞