hi,
i am currently trying to build a distinct count on my cube (mssql2005 analysis services).
But after i added the discount count on the field i want to and start the processing, the following errors appear.
hi,
i am currently trying to build a distinct count on my cube (mssql2005 analysis services).
But after i added the discount count on the field i want to and start the processing, the following errors appear.
Try and see what kind of query Analysis Services sends to the relational database during processing of distinct count measure.
You will see it sending a query containing ORDER BY clause asking relational database to sort results accourding to the distinct count measure.
It it possible the view you defined your partitions on, brings data sorted differently
Any new data becomes avaliable during processing of the partition
The error indicates Analysis Server detecting inconsistencies in sorting of data coming from relational database.
See if you might need to define collation correctly for your sort.
Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
Thanks edward.
as i drew data from Oracle view, the collation must be specify correctly.
check the Oracle collation and discovered it is binary.
changed the distinct count to binary collation and it works.
thanks.
-
HY
Could any one explain about the error and solution elaborately. I am not sure how this can be rectified. Appreciate any help.
Thanks
Hi there,
We encountered the same issue:
And changing the collation to binary allowed us to process the cube ...
But I still don't understand why I got the error with the collation set to SQL_Latin1_CI_AS
By the way, I also encountered a difference of 1 by browsing the cube and when I count on the table:
Browse on the measure with the distinct count = 800
Result of "select count (distinct (sessionid)) from dbo.facttransaction" = 799
(NB: some sessionid are NULL)
Does the cube take in consideration the NULL values
Thanks
OK found why I got the diff :
Analysis Services handle a NULL value like a 0 value in a DISTINCT COUNT measure
I am facing a similar problem shile deploying an analysis services project. Could you please let me know the solution for the same.
Regards
I faced the same problem with fields with different lenghts and odd characters like _ . etc.
I converted the contents of the field to a hash e.g. by using CHECHSUM, and I had no more problems