Problem on using T-Code KEFC

Question

Experts ,
These days, I am trying to insert as many lines as possible to COPA table ( CE1XXXX) in order to perfrom some performance test. However, I meet some problems, after struggling for several days, I come here to beg for help. Any response is appreciated. Here is my records.
1,  I choose EDT to insert data to COPA table, so KEFC is recommended.
2,  I self-made a script tool to generate .csv files with the lines like "aaa;bbb;ccc;...;" and then uploading .csv files through KEFC
3,  Uploading .csv files through KEFC works! but it seems that KEFC doesn't support uploading BIG .csv files.
     According to my experience, it supports less than 10MB-size csv files. (maybe influenced by hardware condition)
4, So I have to upload 10MB-size .csv files one by one through KEFC
5, But when the COPA table becomes bigger, the speed of uploading become slower and slower.  Now a 10MB-size .csv file cost half an hour to be uploaded compared with the orignal 15 minutes.
6, I am trying to set some background jobs to pipeline these uploading works and release ourselves, but the created job always be forced to cancel with the following error:
25.10.2011 16:43:07 Background processing and local files are not allowed                        GD           337          E
7, We guess the above error is caused by the subscreen which jumps out after we clicking "execute" and inquiry users's selection.
  so we want to temporarialy modify the source code of KEFC to avoid that subscreen... but finally we have to give up espencially because it causes headache.
8, Now, we are still struggling to upload .csv files through KEFC one by one with more than half an hour each, and the worse news is the cost cost still increase but we still have hundreds of .csv files to be uploaded.
9, We appreciate any response and help. Thank you very much!
Finally, SAP Network is great, I hope one day I can also be a solution providor. I will : )
Edited by: Fenglei Gu on Oct 26, 2011 4:30 PM

Answer 1

Hi
How will your real time scenario be? Its better if we simulate that
you must estimate like this
a. No of invoices from SD that may be raised
b. Any manual postings from FI or other misc postings from MM
c. Montlly Order settlements to COPA posting variances
d. Any assessment cycle from CO to COPA for marketing overheads - estimate the no of segments you will be distributing upon
Based on this estimate, sit with your basis guy and do the server sizing accordingly.... Use Summarization levels in COPA to improve performance
br, Ajay M