Two fact table with different granularities and star model

Hello 🙂

I am facing an issue on my PBI datamodel.

I have a star-model with 2 fact tables that are linked to the same dimensions

Fact 1 -> Dim A / Dim B / Dim C

Fact 2 -> Dim A / Dim B / Dim C / Dim D

Fact 1 & 2 are EXACTLY the same except that Fact 2 has two additional columns that increase the volume (list of contracts) that are stored in a specific Dim D

Initially the goal was to build only one model on Fact 2 + all dimensions but since the volume is huge the performances are not very good (I work in Direct Query to an Oracle DB).

I was thinking of something like this :

- keeping the 2 models

- when the visuals are not using Dim D then use the measure built on Fact 1

- if a visual using Dim D is used then use the measure built on Fact 2

This way we will keep good performance for all reports except the ones using Dim D (that's OK).

Any way I can achieve this ?

My idea was to create two measures (Meas_Fact_1 and Meas_Fact_2) and then Meas_Fact_3 that will call either Meas_Fact_1 or Meas_Fact_2 depending on Dim D... but I can't figure it out 😞

Thanks for your help