Hi all! I am in the start-up phase of development.
Using SQL Server 2000, SP3.
Analysis Services, SP3.
Standard Edition of Windows Server 2003.
Hardware: 2 18Gb drives, 632 Mb memory, 2.6 GHZ Pentium4
How realistic is it to have a single dimension with approximately 2.5
Million rows? Additionally, I have approx. 20 other dimensions
assigned to the cube. Using Excel XP, performance on the client is
extremely slow. Any thoughts?
Thanks in advance, Jeff.

On 13 Sep 2004 08:19:10 -0700, jnalley (AT) easycare (DOT) com (Jeffrey Scott)
wrote:

A simple rule for this scenario... how many values would your users
expect to browse to select the Dimensional Data they want to filter
by?

Try adding an Active/Inactive flag to the junk data, or perhaps
SnowFlake the data a little so that it's more manageable from a User
perspective.

No user will be interested in selecting more than a handful of
elements from perhaps a few hundred possibilities.

M