Hello,

I use Custom Rollup Formula to calculate the aggregations for one
dimension. The formula is like:

Iif(
Isleaf(Time.CurrentMember),
CalculationPassValue([Measures].CurrentMember, 0),
Iif(
[Measures].CurrentMember.Name="Avg Audience",
Avg(Time.CurrentMember.Children),
Iif(
[Measures].CurrentMember.Name="Cum Audience",
Max(Time.CurrentMember.Children),
CalculationPassValue([Measures].CurrentMember, 0)
)
)
)
and I will have 3 more other measures need to be put into the formula
with different calculations. When browsing the data, it took a while
to load. I am afraid that OLAP was calculating the aggregation when I
browsed the data. I have two questions:

1. How can I force OLAP calculate the aggregated data when processing
the cube?
2. If 1 is impossible, how can I rewrite the Custom formula to improve
the performance when browsing data.

Thanks,

No aggregation will occur above the lowest level where custom rollup
applies, so your best bet may be to use other approaches. See this
earlier thread on the subject:

http://groups.google.com/groups?hl=e...8&edition=us&s
elm=3E8BFEC4.E35C4AE%40dsslab.com
Subject: Re: Missing aggregations in partition
View: Complete Thread (7 articles)
Original Format
Newsgroups: microsoft.public.sqlserver.olap
Date: 2003-04-03 06:29:48 PST

(answers inline)

MrAbeille wrote:

Yes.

Yes, that is exactly the way I would build it.

For when precalculated aggregates can't help you. Actually, I think that
there are very few good uses for custom rollup formulas, and that most
cases I've seen where people use them would be better handled using
calculated members.

Because the agg wizard has no knowledge of which levels have the
formulas and which ones don't.
...

- Deepak

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!