DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20710>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20710

Performance Problems-NEED HELP!





------- Additional Comments From [EMAIL PROTECTED]  2003-06-13 16:17 -------
Hi Andreas,
(sorry that I previously addressed you by your last name)
So the change from 2.4.1 to 2.5.1 gave you noticable but not significant 
improvement, and this is what I expected.  But every 10% (or so) helps.

Recent development changes would probably give a similar improvement yet again, 
but it sounds like you are looking for a 3 or 5 times speedup and I can't see 
how that is going to happen within Xalan.  

Have you tried using using XSLTC?  Given your code you should be able to set 
the Java property:
javax.xml.transform.TransformerFactory=
org.apache.xalan.xsltc.trax.TransformerFactoryImpl

I haven't looked into the cost of creating a Transformer (myTransformer) vs. 
doing the transform. A Transform object is not thread safe, so you can either 
create a new one with each servlet request (which is what I presume you are 
doing)  You can re-use a Transform object, but only in one thread at a time, 
not concurrently. So an alternative to creating new ones is to have a pool of 
Transform objects and have each servlet request get a Transform object, use it 
and give it back to the pool.  I don't know how much this will help, but it 
would be worth a try.

It is hard to accept this as a valid defect, because there doesn't seem to be a 
performance regression in Xalan, just that it isn't fast enough.  However, 
please create attachements of your XML and XSL input and so we can do some 
performance analysis of Xalan with them.

Regards,
Brian Minchau

Reply via email to