The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.
Hello Community -
We are converting our Power BI instance from referencing mostly flat files on sharepoint to a true SQL Server Data Warehouse and I am looking for some best practices on the Power BI architecture.
We have many reports that will reference the same data source/tables from our DW, but may have varying measures, filters etc against the data. We DO NOT want to have several reports that are connected to the the DW but refresh separately or require us to add the measures to each report individually.
Would best practice be to set up a Dataflow(s) and then create one semantic model and DirectQuery that Semantic Model? Or are the other, more efficient solutions that do not require the direct query, because I know there are limitations within that which do not allow us to add calculated columns.
We are beginning this conversion in the next week or so, so any information, articles, best practice tips are greatly (!!!!!!) appreciated.
Thank you as always, Community!
Ryan F
Solved! Go to Solution.
Hi @ryan_b_fiting ,
Yes, using dataflows is a good idea if you are making transformations in Power Query.
Import mode is an alternative to Direct Query, but using it depends on the amount of data you have.
Here is some additional information:
https://radacad.com/power-bi-development-best-practices
https://radacad.com/directquery-live-connection-or-import-data-tough-decision
Hi @ryan_b_fiting ,
Yes, using dataflows is a good idea if you are making transformations in Power Query.
Import mode is an alternative to Direct Query, but using it depends on the amount of data you have.
Here is some additional information:
https://radacad.com/power-bi-development-best-practices
https://radacad.com/directquery-live-connection-or-import-data-tough-decision