[xep-support] Different performance on different systems

[Chris Bowditch]

uwe.goersch at lhsystems.com wrote:

> Hi,
> 
> I am facing the following situation: When XEP renders a FO source into a 
> PDF target, I am facing a bad performance on one system and an 
> acceptable performance on the another system. The two systems are using 
> different hardware, software and network infrastructure. However the FO 
> source, the XEP version 3.8.4 server edition are the same. (Suprisingly) 
> the hardware performance features on the system with the bad performance 
> are better.
> 
> I measured
> - 120 seconds of waiting time for the system with bad performance and
> - 9 seconds of waiting time for the system with acceptable performance 
> for executing the code marked by ***HERE*** in the output (snippet) 
> generated by XEP.
> 
> ...
> (generate [1][2][3][4][5][6][7][8][9][10][11][12][13][14]***HERE***))
> ...
> 
> In order to find out the reason for the different behaviour, I would 
> like to know what the XEP renderer is doing at this point of progress. 
> Did anyone face similar problems or does anyone have a hint how to 
> analyze this problem best?
> 
> Thank you very much in advance!

What sort of environment are you running XEP in? Is it running inside a 
Webapp,  from commandline or an embedded java application? We found XEP 
doesn't scale across multiple CPUS in a multi-threaded java application. 
Others have reported the same on this list. Adding more CPUs actually 
makes the performance worse!!!!!!

Perhaps the machine which you say is superior has 2 CPUs versus 1 for 
the machine which yields better performance.

Chris

-------------------
(*) To unsubscribe, send a message with words 'unsubscribe xep-support'
in the body of the message to majordomo at renderx.com from the address
you are subscribed from.
(*) By using the Service, you expressly agree to these Terms of Service http://www.renderx.com/terms-of-service.html