We have created a few data marts and have many cubes now. We are moving
forward with a Reporting Services endeavor to get a common reporting
platform. We have report servers that basically have dumps of our OLTP
DB's.
a.. Isn't creating a big data warehouse the appropriate place to pull most
reports?
b.. So what really is the difference between using a "data warehouse"
verses "data marts" in creating cubes?
I think we need to have one big data warehouse for our reporrting needs, and
as we create more cubes we create them from the dw - am I wrong here?

Also; I would think that since some of our cubes fit the UDM methodology -
the data isn't in the data warehouse - so that would mean they could only
report on the cubes, right?