Re: History cube design question

#1

As you know adding measures requires Cube full processing.

My idea is that you create Measure dimension explicitly, such as Account
dimension,
because Incremental Updating of dimension dose not require Cube processing.

Finally you'll be able to process the last partition selectively.

Ohjoo

"Ashok" <Ashok (AT) discussions (DOT) microsoft.com> wrote

Quote:

In SSAS 2005 - I have History cube with 2 partitions 2005 and 2006 data.
Both
has different tables in data source view. In cube both partitions are map
to
same measure group, so there is no different measure groups for 2005 and
2006
just to make it clear.

Time to time I need to add new measures in latest cube (2006 partition)
data
those measures don't have data for 2005 partition.

Right now when I add new measures in 2006 I have to full process the 2005
and 2006 both partitions. Is there any way or way to design my cube so
that
when I add new measures to cube I can only full process 2006 partition. If
user selects my new measures with 2005 date they can get null/zero that's
fine.

Thank you for sharing your ideas - Ashok

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Re: History cube design question

microsoft.public.sqlserver.olap microsoft.public.sqlserver.olap

Re: History cube design question - 12-21-2006 , 08:42 PM