on 02-02-2010 7:29 AM
Hi everybody,
we have a java-mapping where a large message has to be mapped. We tested already successfully. In another test we got a error in SMQ2.
So between the first and the second test, there must be any difference in the system. We did'nt change any basic parameters. There was also no traffic on the system in both tests. So we wonder what could be the reason?
Regards
mario
Edited by: Mario Müller on Feb 2, 2010 2:29 AM
hi,
>>>We did'nt change any basic parameters. There was also no traffic on the system in both tests.
hope you know that it's not only the traffic that can slown down processing XI messages...
(it can also be different jobs running on XI or anything else running on server which made a few second difference)
Regards,
Michal Krawczyk
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Mario,
Is it always like that : 1st test is fast, 2nd (and maybe others) are slow ? Are you using SAX or DOM type parser for your mapping ? Maybe full GC running at the same time ? You may have to monitor system/J2EE resources while performing your tests so you can make sure there is nothing else running while mapping is being used ...
Chris
Hi,
GC - garbage collection
>>As the java is a compiled source from a third party I can't say if it is DOM or SAX.
you can always recompile and check
>>>FYI: The java is parsing a flat EDI-message into edixml.
if you bought it then probably you need to contact the vendor and ask - if this vendor has any support
Regards,
Michal Krawczyk
Hi Michal,
just FYI: The mapping is out of support. I wrote my own EDI-Parser and it works pretty well.
But the problem I described is much more complicaded as the JAVA not just only converts EDI-flat to EDIXML; it also validates the incoming EDI an sends a CONTRL.
So we can't switch this on the fly. We got to optimize parameters. But as the mapping already was processed successfully, we are confused, which parameters we should change. Sometimes it is the combination of parameters. And the more parameters are involved the more complcated it gets.
... but you already know! ;.)
Regards Mario
Mario,
How big is the processed msg ?
Regarding the validation of the incoming msg, do you know how it works ? Is it using stanard parser validation or is it a custom implementation ? It would be helpful to find out where the performance issue is : mapping ? validation ? memory ? other ?
Rgds
Chris
Hi all,
thanks for your replies. Unfortunytely I received some wrong information. The JAVA was processed successfully. We got an error due to an JCo-failure.
Sorry for the inconvience.
Regards Mario
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
94 | |
11 | |
11 | |
10 | |
9 | |
8 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.