cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
Subject Request to transfer bug to Cocoon: Re: [PATCH] Re: [C2][Xalan2] Xalan2J problems under heavy load using Apache JMeter
Date Fri, 16 Mar 2001 05:32:23 GMT

I believe I now have enough solid evidence that this is a cocoon bug to
transfer the bug report over to you guys.

In my test I am using http://localhost:8080/cocoon/sites/
An exception is thrown pretty quickly on a null ResultTreeHandler from
within ElemLiteralResult.  I did a test in our SourceTreeHandler
(implements the TransformerHandler interface), and checked for reentry to
the startDocument event.  If I can determine reentry to startDocument
before endDocument occurs, I did a stack dump.  This led me to

My first suspect was that the contentHandler member variable wasn't being
properly cleared.  But I would think that would cause all sorts of problems
beyond a thread-based overwrite.  I tried synchronizing that block with the
ContentHandler, and that actually caused the problems to get much worse...
exceptions were being thrown by Xerces like crazy.  I conclude from this
that the same parser object is being used by multiple threads.

So I synchronized with the parser object immediatly after the
(Parser)this.manager.lookup(Roles.PARSER) lookup.  This fixed it!

So my conclusion is that somehow either the pooling of the parser, or
perhaps the pooling of the FileGenerator, or both, is failing.

Is it OK if I transfer, and perhaps, and let you guys
take it from here?  Someone more familiar with Cocoon and Avalon internals
can diagnose the problem from here far more efficiently than I can.


To unsubscribe, e-mail:
For additional commands, email:

View raw message