Hello,
I have a MDX query like this:

WITH
SET [SelectedStations] As
'{[Station].[5433014],[Station].[5433015],[Station].[5433024],[Station].[5433025],[Station].[5433042]}'
MEMBER [Station].[Permited] as 'Sum([SelectedStations])'
SELECT
NON EMPTY {[Year].[2004].[Days].[all days].children} ON COLUMNS,
NON EMPTY {[Seller].[all seller].children} ON ROWS
FROM
[UPPOK]
WHERE
( [Station].[Permited])

without the where clausule is quick, with where clausule is slow.
Is there some way to improve this?

Some facts:
The number of {[Seller].[all seller].children} is 6000
the number of NON EMPTY {[Seller].[all seller].children} is cca. 30
the number of [Station].members is cca 200

the number of [Year].[2004].[Days].[all days].children is 365
the number of NON EMPTY [Year].[2004].[Days].[all days].children is cca. 20

The set [SelectedStations] is for each query different.

Based on the dimension data you provided, the results are sparse, so
using NonEmptyCrossJoin() may help, like:

SET [SelectedStations] As
'{[Station].[5433014],[Station].[5433015],[Station].[5433024],[Station].
[5433025],[Station].[5433042]}'
MEMBER [Station].[Permited] as 'Sum([SelectedStations])'
SELECT
NonEmptyCrossJoin([Year].[2004].[Days].[all days].children,
[SelectedStations], 1) ON COLUMNS,
NonEmtpyCrossJoin([Seller].[all seller].children,[SelectedStations], 1)
ON ROWS
FROM
[UPPOK]
WHERE
([Station].[Permited])


- Deepak

Deepak Puri
Microsoft MVP - SQL Server

*** Sent via Developersdex http://www.developersdex.com ***

Thank you, it's faster.
But I still have a problem. Time grows exponencialy with regard to size of
[SelectedStations] and problem is in WHERE clausule. Is there a solution?
Ondrej Srubar

"Deepak Puri" wrote: