I am a newbie on OLAP cube. What is the right approach when building an OLAP
cube? For example I require 10 fields (attributes) as dimension from a table
in an OLAP cube, should I build 10 seperate dimensions, one for each
attribute or build one dimension with all 10 attributes on it?

--
Message posted via http://www.sqlmonster.com

do you use AS2000 or2005?
what are these fields?
if they are complementary (like customer, customer gender, customer
country...) and can be setup to create an hierarchy then put them into a
dimension

if these fields are more deisgn for cross analysis then create separate
dimensions
(like time and customer name are not common attributes)

generally I group the attributes of an entity into 1 dimension (like
customer dimension contain all customer attributes same for product,
time...)
design this from a logical and easy to use point of view.

"pmak via SQLMonster.com" <u25043@uwe> wrote