Recently I was reading an article by Boyan Penev on DSV’s called “Using the DSV to its full potential” and it got me to thinking about how I use DSV’s…
In SSAS 2005 when DSV’s were introduced I found myself using named queries quite a bit. But the more I read about design methodologies for cubes, I felt like it was better if I used views in the database to actually control what data was being pulled into the cube rather than use named queries. That way I could have better control over the logic and not have to make changes to the cube all the time. In fact the only way that I can see myself using named queries was if I were going to have two fact tables built off of the same view or table. In this case I may use a named queries to filter the data to create the two fact tables. Other than that, I cannot see why I would ever use a named query.
This function of the DSV I do use. Specifically for the “hanger” dimensions for time calculations.
I think Boyan wrote a great article and I suggest reading it. I just can’t ever see myself doing most of the things he outlined in his article simply based on the fact that I can do them all in a database level view.
Do you use the DSV in a the way that Boyan describes?