Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

The FabCon + SQLCon recap series starts April 14th at 8am Pacific. If you’re tracking where AI is going inside Fabric, this first session is a can't miss. Register now

Reply
Anonymous
Not applicable

One giant data model?

Hello Community  -  Having a debate with our sysadmin and would appreciate any guidance.  

 

We have an ERP system that we cannot connect to directly  (so we use dataflows connected to ERP).   We use this to create a large "ERP" dataset.    We build reports from this, both in the desktop, and in the service.     That dataset is in import mode.  

 

We are contemplating breaking up that dataset into subject matter areas datasets.   For example orders and shipments in one dataset.   Then have RMA (returns) and quality data in another data set.   Logistics dataset, etc.    The rationale is that there are unique calculated columns and measures that are specific to certain business logic and may not have relevance to other business units.   And from a report creation standpoint we do not want every report created (in desktop) to have this large model in the background.  We'd rather have that report pinging a smaller dataset rather than the one big dataset.  

 

Comments?  Suggestions?    Does the idea of discrete subject matter datasets make sense?

 

 

2 REPLIES 2
AbhiSSRS
Solution Sage
Solution Sage

This is a perfect ground for an excellent Power BI dataset organization! I recommend you go ahead with this . The approach generally taken is that the power BI datasets are built and deployed to service without any visual in them. So these serve as dataset only reports. These datasets can be then used to build multiple reports by connecting live to them by multiple developers at the same time. This has several benefits :

The updates to datasets done would reflect on all live connected reports automatically.

The visual reports are lightweight and easy to maintain and work with in respective Desktops.

The deployments to environments simply needs Datasource updates in pbix

The self-service goal is better realized where business can even build their own reports connecting to these well defined datasets.

You would be able to use Tabular editor and also manage RLS centrally in the datasets

 

Please let me know if you need further information in this set-up

 

Anonymous
Not applicable

@AbhiSSRS    Great feedback.   Just to confirm, you are advocating that I proceed with the subject matter dataset architecture?    

 

And one related question.   When we reports in the service, is there a trick to being able to start with a blank page that already has a custom color theme applied?   We publish most of our reports from desktop today so we can easily apply a theme.  

Helpful resources

Announcements
New to Fabric survey Carousel

New to Fabric Survey

If you have recently started exploring Fabric, we'd love to hear how it's going. Your feedback can help with product improvements.

Power BI DataViz World Championships carousel

Power BI DataViz World Championships - June 2026

A new Power BI DataViz World Championship is coming this June! Don't miss out on submitting your entry.

Join our Fabric User Panel

Join our Fabric User Panel

Share feedback directly with Fabric product managers, participate in targeted research studies and influence the Fabric roadmap.

March Power BI Update Carousel

Power BI Community Update - March 2026

Check out the March 2026 Power BI update to learn about new features.