[
https://issues.apache.org/jira/browse/WSCOMMONS-556?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12918517#action_12918517
]
Kasun Gajasinghe commented on WSCOMMONS-556:
--------------------------------------------
This is just a small improvement we did as part of several big improvements. We
saw huge use of Strings which took in memory, so we optimized the Strings usage
in places we saw as important. This change is just a small part of it. And, we
can't really say the improvement level only by this change, as there were
series of changes done even in Strings. But I think the difference is less, as
I said in a previous comment.
So, it's really not a big deal. I understand what you are saying You may remove
that if it's unnecessary.
> Performance improvement for axiom-dom module
> --------------------------------------------
>
> Key: WSCOMMONS-556
> URL: https://issues.apache.org/jira/browse/WSCOMMONS-556
> Project: WS-Commons
> Issue Type: Improvement
> Components: AXIOM
> Affects Versions: Axiom 1.2.9
> Environment: Java 1.5
> Reporter: Kasun Gajasinghe
> Assignee: Nandana Mihindukulasooriya
> Attachments: axiom-dom-556-patch2.diff, axiom-perf-improvement.diff,
> FINAL-times_for_large_samples_with_changes.png,
> FINAL-times_for_small_samples_with_changes.png
>
>
> This article published [0] on April 2010, shows that Axis2 has some major
> performance issues compared to Metro and CXF. An analysis suggested that
> most of the problems were with the axiom-dom module's methods, along with
> Rampart. So, first we went on improving the axiom-dom module. Here's a patch
> for the improvements we did.
> Final assessments suggest considerable level improvement. Graphs for a
> sample run of requests with small and large samples will be added to the
> files section.
> [0] http://www.ibm.com/developerworks/java/library/j-jws14/index.html
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
