Starting December 3, join live sessions with database experts and the Microsoft product team to learn just how easy it is to get started
Learn moreShape the future of the Fabric Community! Your insights matter. That’s why we created a quick survey to learn about your experience finding answers to technical questions. Take survey.
I have a single dataset that populates a bar chart based on groups. That bar chart then drills through to a details page to see each individual record.
How can I set up RLS so it limits only the drill through page to allow users to only see their groups but still see all groups on the main bar chart?
Will I need to create duplicate datasets and apply the RLS logic on the drillthrough dataset only?
Solved! Go to Solution.
My solution:
1) Import original dataset, call it XXX
2) In modeling tab, create a new table and set it equal to my original dataset, call it XXX_Secure. (XXX_Secure = XXX)
3) Create a relationship between XXX and XXX_Secure on group name.
4) Configure Row Level Security on the USERPRINCIPALNAME on table Y.
5) Create relationship from table Y to XXX_Security on the group name.
6) Create bar chart off of XXX
7) Create drill through detail page, include all fields you need from XXX
😎 Also add the group name from XXX_Secure to the detail page and minimize the column so it can't be seen.
This will allow all users to see all the data in the bar chart regardless of group name. Every bar in the chart will have a drill through option, but clicking on them will return empty details except for the groups the person viewing the report has access to.
Hi @TCavins ,
RLS is applied at a data-level. It has nothing to do with the tabs on your report.
It will apply filtering to the rows in youe dataset and further on the whole of your report, ir-respective of; if it is your main tab or your drill-through tab.
Thanks,
Pragati
@Pragati11 Hit the nail on the head.
The only thing I would add is, you can use the "View as" button to test RLS before publishing. This will allow you to test your roles to ensure they are working correctly.
My solution:
1) Import original dataset, call it XXX
2) In modeling tab, create a new table and set it equal to my original dataset, call it XXX_Secure. (XXX_Secure = XXX)
3) Create a relationship between XXX and XXX_Secure on group name.
4) Configure Row Level Security on the USERPRINCIPALNAME on table Y.
5) Create relationship from table Y to XXX_Security on the group name.
6) Create bar chart off of XXX
7) Create drill through detail page, include all fields you need from XXX
😎 Also add the group name from XXX_Secure to the detail page and minimize the column so it can't be seen.
This will allow all users to see all the data in the bar chart regardless of group name. Every bar in the chart will have a drill through option, but clicking on them will return empty details except for the groups the person viewing the report has access to.
Can you please make the reference more easy to understand.
I have a similar situation - trying to figure out what is Table Y - for the userprinciplaname().
Thanks again!
It's the table that lists out a record for every login and permission.
Thanks TCavins.
so a table with user name, upn, and Y or N for drill thru permission?
Not a Y/N for permission but actual value. bob@company.com, 'Microsoft' can see all records where column x is Microsoft.
In your Power BI report, you set up the security on the report Under Modeling > Manage Roles and create a role, select the table that has your security (Table Y) and a DAX expression to say [Username] = USERPRINCIPALNAME().
You then set up your model to say TableY links to Companies(M*M) and Companies is linked to your dataset(1*M). This way, the security table filters the companies table that then filters your dataset.
In my Microsft example, security table will have bob and Microsoft. The companies table will have a list of distinct companies in your dataset. Your dataset should contain the company column for each record.
In the relationships, you set Table Y filters Company table. Company table filters your dataset.
User | Count |
---|---|
93 | |
90 | |
90 | |
81 | |
49 |
User | Count |
---|---|
156 | |
145 | |
104 | |
72 | |
55 |