Have you tried profiling the build to see what is actually happening? Just because you are getting an OutOfMemory error doesn't mean that you are running out of heap space. There are a couple of free tools out there including one by Sun http://developers.sun.com/dev/coolstuff/jvmstat (the one that ships with JDK 1.5 is pretty nice). The Sun memory profiler will let you look at others areas besides heap, so I would recommend starting there.

Dion Gillard wrote:
On Thu, 17 Jun 2004 13:24:56 +0100, Matt Read <[EMAIL PROTECTED]> wrote:


Are the unit tests allocating large amounts of memory?

Define large... with a forked JVM for the unit tests I observe an additional JVM footprint of around 16mb (the default min heap size?) as the test execute (JWebUnit tests incidently) the memory usage increases to around 200mb which is well within the range of my machine and max heap size settings. It's possible that JWebUnit itself has a memory leak however I'm 90% sure I can reproduce the problem without running the tests - please tell me if you think it would be useful to do this and tell me what data you'd like me to gather when I do it.


Not just yet, but it is interesting.


3. I seem to have temporarily resolved my issue by doing

the following

but my susispicion is that it will come back as my codebase grows.
   a) disabling all JDepend reports (I've had lots of

problems with

OutOfMemory errors and Jdepend runnning from Ant in the past)
   b) forking the JVM while the Junit tests run
   c) separating removing any other goals from the build

so that I'm

only running multiproject:site.

Does turning any of the above back 'on' cause the failure again?

Yes, the failure is reproducable, would it be useful for you to know exactly which if these options might trigger it?


Definitely. If it's JDepend helping out, or the junit tests would be useful.


4. At present the multiproject is made up of only 3 sub-projects, I
was intending for this to grow to approximately 20 over the

next week

but will now have to rethink.
5. I can't really clarify the "exact failure" any better without
knowing what kind of information you are after?

Generally looking for patterns about where it's failing. e.g. always during xdoc or always during XXX.

The majority of the time during xdoc, but not at any determinable stage within the xdoc goals, sometimes during transforms, sometimes while building the aggregated menu.


FWIW, I have a multiproject of around 20 projects running in less
memory than what you've defined here. How many source files are we
talking about?

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



--
Eric W. Hauser
Application Developer
accessIndiana..."linking hoosiers to government"
http://www.IN.gov
10 W. Market St., Suite 600
Indianapolis, IN 46204
Phone: (317) 233-4007
Fax: (317) 233-2011

**********************************************************************
CONFIDENTIALITY NOTICE:
This E-mail and any attachments are confidential. If you are not the intended recipient, you do not have permission to disclose, copy, distribute, or open any attachments. If you have received this E-mail
in error, please notify us immediately by returning it to the sender and delete this copy from your system.
Thank you.
accessIndiana, MyLocal.IN.gov, CivicNet
**********************************************************************



--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to