Forum Discussion
Row Level Security error for admin/author
I've authored and uploaded a dataset to a workspace I am an admin for.
When I try to connect to this Power BI Dataset in desktop (both Nov and Dec versions), it's giving me an error when building visuals:
Error Message:
The query encountered security filters on a table relationship that cannot be reconciled for the current user. Please contact your administrator or developer to review role memberships and security definitions.
I have a DAX expression on a table with user emails from Azure Active Directory: [DomainName] = USERPRINCIPALNAME () and I have the Apply Security in Both Directions checkbox ticked. I'm not in this table. When the RLS is removed or the Apply Secuirty is unchecked then the error stops but that would allow users to see all the data so isn't an option.
This was working before the November release and the support team I'm talking to at Mindtree don't seem to know what is causing it and suggesting I add myself to the RLS in Power BI Service.
Could someone confirm if things have changed in these updates and I no longer have unhindered open access as an author or admin for this dataset. I am unsure how to create RLS under manage roles that will let me see everything.
So this is a bug. It's not going to be fixed until Mid-March. There is a workaround that doesn't last long where you have to open the dataset in the powerbi service and then connect to it via the desktop to create your reports. Last a few hours then stops. Absolute madness this has been like this since November. Five months for a fix so I can do my job. Well done Microsoft.
8 Replies
Hi Anonymous
I don't think the Version upadates have some to do with the RLS, maybe resetting the RLS could solve the issue.
Add your user name to the Table and then give yourself permission as an Admin of the report.
https://docs.microsoft.com/en-us/power-bi/admin/service-admin-rls
https://www.youtube.com/watch?v=9wN33rTaiB4
Let me know if you needed more help
So steps taken:
Removed RLS
Added myself to RLS table
Re-added RLS with DAX as above
Uploaded dataset to workspace where I'm an admin
Result:
It's also blank when I uncheck Apply Security both directions. Conclusion: Adding my email to the RLS table and giving myself admin rights in the workspace doesn't allow me to see the data at all.
So, being this RLS set up worked up until November, something must have been changed. I've not made any changes to the dataset security/RLS-wise. I will have to revert to a previous version to see if I can author that way successfully.
Hi Anonymous ，
Would you please show us the formula you used in configuring RLS in Power BI desktop? As far as I know, if you are the admin of the workspace, the rls will not work for the admin user. Rls will only work for users with view permission only.
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
Best Regards,
Dedmon Dai
So this is a bug. It's not going to be fixed until Mid-March. There is a workaround that doesn't last long where you have to open the dataset in the powerbi service and then connect to it via the desktop to create your reports. Last a few hours then stops. Absolute madness this has been like this since November. Five months for a fix so I can do my job. Well done Microsoft.
Hi,
I was having the exact same issue.
The formula for configuring RLS in PBI desktop is the following
And I've also made a 'no RLS' group:
And had checked the Both ways security checkbox.
Also, when connecting to the published dataset from desktop, still was not able to see any data.
What fixed the problem for me was
- Deleting the dataset from the workspace
- Unchecking, and checking back in the both way security option
- Publishing the dataset back to the workspace
If I did it by overwriting an already existing dataset, then it did not solve anything.
Maybe this could also work for you or anyone else experiencing this bug.
I can't delete my dataset. It is the only one we have for all our reports sof we remove it we remove all our reports and we have lots of them. I thought that was the whole point of this.....you should keep your datasets to a minimum and author all your desktop reports off of one. It just means writing reports takes way longer than it should and hasecome annoying. It's a major fail on their part and they aren't even mentioning it in the current support issues.
