First of all, thanks for all the great responses. This is a great board!

I'm trying to break my dataset into multiple cubes across multiple machines.
Basically, this year's data on one machine, last year's on another, with a
third as the host for the virtual cube. All cubes have the same measures
and dimensions.

I've managed to build each cube on it's respective machine, i've linked the
two cubes together in a virtual cube on the third machine, but there's a
problem (at least one!): I use a dimension representing time and since they
are named the same for each cube, I can only select one during the virtual
cube creation. Depending on which one I pick, that's the year's data that
is visible. My question is: is there any way to have a dimension that
would represent the span of the dates in the data and have the ability to
drill using that dimension?


Thank you,
Malcolm

I understand all cubes have the same measures and dimensions.

It means that you have to use partitioning instead of creating separate
cubes. You can save partitions on local or remote machine. For example,
current year partition on local machine, last year partition on another
machine (remote partition).

Ohjoo Kwon


"Malcolm Toon" <mtoon@ no-spam.retailsolutions.com> wrote