How to set PDP effectively

Try to explain this as clear as I can: this is a simplified version, I have more data columns need to be considered.


I have a dataset A.  I created 3 different dashboards using this dataset A.   

DashboardNameData columns
all webpagesDAwebpage region, web business area
campaignsDBcampaign region, campaign business area
a group of webpagesDC 


User groupsusers


User group1, 2 have access to Dashboard DA, DB and user group 3 have access to DC.   

1. User group 3 who are also in group1 or 2 should have full access to all data for dashboard DC.

2. User group 1, 2 should have access to DA based on web region and web business area;

3. User group 1, 2 should have access to DB based on campaign region and campaign business area;


I am not sure how to handle this.  People in the same group are seeing different numbers from the same report because of the PDP setup.  


Thank you.







Best Answer


  • I'm 95% confident that PDP permissions are ADDITIVE (so if i'm I'm in group 1 and group 3 i will see the data available to group 1 AND the data available to group 3).


    It sounds like you want to use PDP to provide filtered views of your dashboards and that's NOT the same thing as providing filtered sets of Data.  assume that if you give me access to data via PDP i will ALWAYS see ALL the data (regardless how you want to filter my cards.


    That said for dataset you could add binary columns to your dataset, ForDashboard1, ForDashboard2, ForDashboard3 then use that column to hard code filters into your dashboards and subset the data with respect to whether you want it shown in the dashboard, but not necessarily tied to a PDP policy.

    Jae Wilson
    Check out my 🎥 Domo Training YouTube Channel 👨‍💻

    **Say "Thanks" by clicking the ❤️ in the post that helped you.
    **Please mark the post that solves your problem by clicking on "Accept as Solution"
  • @jaeW_at_Onyx thank you for the explanation.  That is probably why they see different numbers. 


    Creating binary variables in the dataflow is tough for me:  because this is a huge dataset with 114 M rows, and we had time-out issue before.  So I changed the whole process by running 2018, 2019 dataflows separately (just need once) and stack them using assembler, now I only need to run 2020 data daily and stack them with the historical data. this saved me tons of running time.  Therefore, it is not easy for me to modify the code.   In addition, I will create more dashboards from this dataset, it will be very challenge to keep adding new binary variables.   BTW, I still have not got dataview beta, my CSM said it is closed.


    Are there other ways I can do to prefilter out for each group?  I am getting so many question recently.


    Thank you.









This discussion has been closed.