Dear all

I know that SSAS creates aggregations depending on attribute relationships.
This works fine.

It happend now that business users are grouping dimension members in a way
that was not planned and I have no attribute for it.

I first thougth of making a named set in SSAS, but this will not make any
aggregations, right?. I think I have to introduce a new attribute and create
a attribute relationship to force SSAS to create aggregations.

Is that a good approach?

Thanks for your help,

Regards, Marc


But I have now the case that in some members dimensions I have a

in AS2005 the caching capability of the server allow you to create sets and
the server will keep the results in cache.
but the server can't preaggregate the result of the set.

so, try using the set option if its the easy solution for you and see the
performance results.
adding an attribute is good too, but raise the storage required; and this
required a fullprocess if you want to create new set of members, while sets
defined in the cube can be defined at any time without a reprocess.


"Marc" <Roger_Rombooth (AT) community (DOT) nospam> wrote