Skip to main content
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

Did you hear? There's a new SQL AI Developer certification (DP-800). Start preparing now and be one of the first to get certified. Register now

KenVraagaard

Materialized Lake Views (MLV) IDENTITY insert and MERGE would enable slowly changing dimensions

Submitted by on ‎03-19-2026 02:57 PM

Two features are missing in order to use MLVs to create a dimensional model: Identity insert and Merge. Cf. Databrics Lakeflow.

 

MLV could be a game changer for data warehousing with these features included. Without solution is only partial and now relies on additional steps (workarounds) to create actual SCD Type 1/2 dimensions and then (re-)refresh MLV fact tables relying on those dimensions. Fabric team, you are only half way there with MLV - please don't stop yet.

 

Also, running into error when trying to create MLVs with spaces in column name, e.g. `Customer Name`. Would also require significant workarounds to use MLVs with Direct lake Power BI semantic models without proper column naming allowed.

Status: New
See more ideas labeled with:
Comment