Forum Discussion

TalAruety's avatar
TalAruety
Icon for Advocate I rankAdvocate I
8 years ago

bypass role level security for certain measure

Hi

I have implemented Row Level Security (RLS) in my model and cant figure out how to achieve partail filteration, let me explain:

My data model collects data from different farmers, each has diffrent ID (Field named "MR")

All the visuals in this report presents the same measure with diffrent filter context.

 

The data in visuals 3,4 shows the measure diced by Employye and month

The data in Visual 2 shows the value for all the rows inside the selected farmer ("MR")

the data in Visual 1 shows the value for all rows in model.

 

Each farmer is defined as role in RLS so he can see only his own data

I want that to apply to visuals 2,3,4.

Visual 1 howevers is provided in order to give a benchmark regarding how am i cimparison with the other farmers in the data model.

 

So if i have farmers 11,12,13

when farmer 11 is loged it, he should see only his data in visual 2,3,4 as well as in the slicer for "MR"

but the value for the entire dataset for all the farmers in visual 1.

 

As RLS is implemented on the data set, even if i write a different meaure which cancels filter context on the "MR" dimension column, still RLS will not allow farmer 11 to see the value that includes farmers 12 and 13 in visual 1.

 

Is there a way to implement this requirement?

 

 

Need Help

2 Replies

  • v-huizhn-msft's avatar
    v-huizhn-msft
    Icon for Microsoft Employee rankMicrosoft Employee
    8 years ago

    Hi TalAruety,

    Row-level security (RLS) restrict data at the row level of a dataset. If my understanding is right, you create visual1, visual2, visual3 and visual4 using one dataset collecting data from different farmers. So you if we create a role for the dataset, all the data in it will be effected, we can not bypass role level security for certain measure, more details, please see: Row-level security (RLS) with Power BI.

    Best Regards,
    Angelia

    • TalAruety's avatar
      TalAruety
      Icon for Advocate I rankAdvocate I
      8 years ago

      Hi v-huizhn-msft , Thank you for the reply!

      Are you familiar with a patern where i can achieve this goal?

      In this case it is useful for the farmer to know how has he performed with comparison to his peers.

      I see 3 possible options:

      1. Duplicate my Fact table, without linking it to the "MR" dimenision which filters the original data, and calcluate the measure from it
      2. create a new data set, create a dashboard and put elements from the filterd and the non filterd data in a single dashboard.

      First option should achieve that, but then inflate my model and need to maintain redundunt set of meaures

      Second option can work but as dashboards are static i have to generate a representation for each scenario, ending with ton of dashboards.

       

      Any better suggestions?