Snowflaking for two things,

Hi

Currently my proposed dimensional models are very much star-schema.
Except for 2 dimensions ....

DimAudit and DimSource.

Since all my dimensions and facts will be referencing DimAudit and DimSource it will result in a bit of snowflaking.

Is this okay?

Or should I rather have a field in all my tables called: SourceCode which will contain a text code/name identifying the source of the data records?

(I don't want to start snowflaking again like I did in my proof of concept several months ago)

I assume dimAudit is a run log and you tag rows when updated or inserted by the load process. Having a FK on every row (dimension or fact) is fine. I tend to consider such tables as 'back room' tables that don't play an active role in the end-user experience.

As for source code, I use it as part of the dimension's natural key. This allows you to integrate data from different sources and ensure unique natural key values. The source code may represent a source system or it may represent an authority that is the basis for the particular code. I would also have a source code table, but again, this would be a 'back room' table purely for documentation purposes that contains a list of codes and descriptions. There would be no need to actually define an FK reference as such codes are usually set by literals in the ETL code.