SQL Views | Do you import fact table with a bunch of joins? Or seperate fact and dim tables?

[rzavgazaryan]

Hi all -

When you're connecting to a SQL db, are you usually creating some sort of view in the db to connect to? In my experience, I've usually just make 1 view that contains all fields and dimensions required. But the issue with this, is that when I import it, it's not in a star schema model.

My question is, do folks usually seperatly import dimension tables? and have their fact table as a stand alone? When I was getting into Power BI, I would create star schema's from a flat CSV file by creating indexes, linking PK/FK together in the data model view, etc.

Looking for some best practices here - thank you!

[AlexisOlson]

In general, I try to have views or tables shaped like I want on the SQL server so that I only have to load them without many additional steps in the query editor. This means loading separate fact and dimension tables rather than a big Frankentable that needs to be separated into pieces.