How to handle large CO-PA extraction problem challenges

Question: I am implementing BW it is short of a nightmare our PA system is pretty large about 60 million records in the CE1 table. When I run a PA extract it takes a longtime and then dumps with PSAPTEMP space issues. My bloody dumb BASIS person like everywhere tells me that PSAPTEMP on R/3 is 30GB and my SAP extractor is crap. I am bewildered when I run the extract for 1 period event that takes a long time and atleast works if I go period by period it is longtime for me to extract the data.

Can any of you please throw some light on some sensible CO-PA extraction tuning/optimisation techniques please.

Many thanks in advance.


Anita

Answer:
hmmm... never had a real problem (but then again we only have 4.5 mio records)
actually, now that I think of it... we did have some tablespace problems when we first started (3 years ago). problem was PSA table... are you loading through PSA? if you are, try running it without loading the PSA for a change and see whether you get the same problems.
I believe there are some OSS notes out there to "split" up tablespaces... might be worthwile looking into in your case.
another question: are you doing one load only??? 'cause in that case you will get problems... try splitting them up (per fiscal year for example).

hope this helps,

El Belgio
Copyright ?2007 - 2008 www.jt77.com