Performance problems with SQL data in Transformer.

[Recce]

Hi,

I am reading several tables from a SQL database in our company. When Impromptu join the tables and extract the data then the performance is good however, when Transformer extracts the data via the .iqd then it takes forever to extract and transform. I have even selected the 'Maximise data extraction option'.Does anyone have any suggestions in terms of setting changes or anything else that I should look at in order to solve the problem. Thanks

 

[DoubleD]

Are you running the IMR and the cube build on the same PC?
Look at the cube build log to verify that it is the database time that's taking longer, versus the cube build itself.
From a performance standpoint, ensure that you are ordering and grouping your columns based on cardinality.
Also ensure that your functions are all processing on the database and not on your desktop.

 

[Recce]

Hi DoubleD,

thanks for the reply. Yes, I am running the IMR and Cube on the same server / PC. I will look at the log to see what could be the problem. I only have a Date dimension in the cube as far as calc's is concerned. The rest of teh fields are simply extracted and transformed into their dimentions. I will let you know, thanks. Feel free to keep making suggestions if you think of anything.