January 28, 2004 at 5:58 pm
How do we decide which schema to pickup for your dimensions? The only difference is the physical implementation. Snowflake schema more normalized. Star schema is easier for direct access. Is this the only difference? If so why don’t we MS take Snowflake totally out of Analysis services?
Shas3
January 29, 2004 at 8:13 am
There are trade-offs with both types of schema. Here is some more info from Microsoft:
There's also a reach article about why snowflake makes a good data warehouse design. Haven't reviewed it in depth and it's heavy on the math side, but here you go:
http://www.dcs.bbk.ac.uk/~mark/download/star.pdf
K. Brian Kelley
@kbriankelley
January 29, 2004 at 10:32 am
Oh this is what i am looking for. Thanks Brian
Shas3
February 3, 2004 at 9:34 am
The simple answer in regards to AS is that either one works fine as the data is loaded into the cube structure.
The general rule I follow is if you need to query the relational data, star schemas generally have better performance. If you do not, it is easier to manage the data loads in a snowflake, so I use snowflakes if I am only concerned about the MOLAP data.
Steve Hughes, Magenic Technologies
February 3, 2004 at 11:01 am
I would add that a lot of the physical questions depend on your SCD's and how you handle them. I reccomend any articles by Mr. Ralph Kimball for theory and design questions.
HTH
------------
Ray Higdon MCSE, MCDBA, CCNA
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply