Hello all,
I have a problem concerning differences in record counts between the measure group and the fact table.
for debugging such a case in Analysis Services 2000, I would have run process on the cube, copy the SQL statement that was generated and debug it using Query Analyser. in sql 2005, I tried doing that but got the sql statement without the joins to the dimensions so no much help in that...
does anyone have a suggestion on how to debug such a case?
Thanks,
Momo
In SSAS 2005, the fact table is not joined to the dimensions during processing unless you implemented a reference dimension and opted to materialize it. In this situation, you will see a join to the intermediate dimension table.
B.
|||
Bryan C. Smith wrote:
In SSAS 2005, the fact table is not joined to the dimensions during processing unless you implemented a reference dimension and opted to materialize it. In this situation, you will see a join to the intermediate dimension table.
B.
So can anyone explain why the poster is observing dropped rows?
In my case I am finding the cubes are not showing any financials. Fact tables are populated correctly. Dimensions appear to be joined correctly - but I understand this wouldnt matter anyway?
Is there a way to see how the cube process is getting its data? Is there a way to check the integrity of a data source view? What is the consequence of incorrectly joining tables in your data source view?
Debugging in AS 2000 seemed much easier!
|||
When you process the cube, you have access to the queries submitted. Take a look at those and test the row counts for each.
B.,
No comments:
Post a Comment