Hi - I'm new to AS. Using VS 2005 AS Project with SQL server 2005
developer and AS running on my workstation. My incoming data is set up
as monthly inputs in the form (date, testnumber, dept, manager, cost,
hours). It's journal info with no index. My first inclination when
building a cube was to put all the monthy tables in a data source view
and try to join them somehow. unforntunately it wasn't immediately
intuitive so I couldn't figure it out. I'm tempted to load the data
into a single large table and be done with it. Is these the easiest
thing to do or is there some magic I'm missing? I don't want to build
a new dimension every month. I'd like to just add the table and
reprocess the cube. possible.

thanks

bill

If your monthly input is big, you can setup a partition for each month and
specify that table as a source for that partition. If it not that big or you
don't have Enterprise Edition you can just setup a named query that is going
to select data from all those monthly tables and do UNION ALL between them.

If your date is a datetime you can create a server time dimension and use
that for your time dimension.

Boris.

"bill" <oddworks (AT) cox (DOT) net> wrote