Reasons for perfomance decline - ideas wanted

[Thorsten Scherler <th...@juntadeandalucia.es>]

Hi all,

I am trying to find out why one of my projects suffer perfomance
declines.

I opened an issue around the problem
https://issues.apache.org/jira/browse/FOR-1157 and added some testing
file to
https://svn.apache.org/repos/asf/forrest/branches/dispatcher_rewrite/FOR-1157

Both wars are basically the same but have some minor difference, mainly
the dispatcher library and its dependencies. However I debugged both
versions and tested with different tests. The most important match I am
debugging is the following:

<map:match pattern="**.html"> 
 <map:generate src="lm://resolve.structurer.{1}" type="jx"> 
  <map:parameter name="lenient-xpath" value="true"/> 
  <map:parameter name="getRequest" value="{1}"/> 
  <map:parameter name="contextPath" value="{request:contextPath}"/> 
  <map:parameter name="getRequestExtension" value="html"/> 
 </map:generate> 
<!-- 
  1) <map:serialize type="xml"/> 
--> 
 <map:transform type="dispatcher"> 
  <map:parameter name="cacheKey" value="{0}"/> 
  <map:parameter name="validityFile"
value="lm://resolve.structurer.{1}"/> 
  <map:parameter name="request" value="{1}"/> 
  <map:parameter name="type" value="html"/> 
 </map:transform> 
<!-- 
  2) <map:serialize type="xhtml"/> 
--> 
 <map:transform src="lm://hooks-to-html.xsl"/> 
<!-- 
  3) <map:serialize type="xhtml"/> 
--> 
 <map:transform src="lm://namespace-stripped"/> 
<!-- 
  4) <map:serialize type="xhtml"/> 
--> 
 <map:transform src="lm://strip-dispatcher-remains-html.xsl"/> 
 <map:serialize type="xhtml"/> 
</map:match>

In one test I cut the pipeline after the generator (1) and the result is
surprising. Since there is NO difference between the old and new version
in the generator I was awaiting that the performance should be the same.

But 55,5 (new) vs. 64,5 (old) are not even close to be the same. I do
not see any reason for performance penalty since I am not actually
testing my development but a general component that did not changed at
all.

Has anybody an idea what the reason for the performance decline can be?

salu2
-- 
Thorsten Scherler <thorsten.at.apache.org>
Open Source Java <consulting, training and solutions>

Sociedad Andaluza para el Desarrollo de la Sociedad 
de la Información, S.A.U. (SADESI)