[ https://jira.jboss.org/jira/browse/JBAS-7248?page=com.atlassian.jira.plug... ] Brian Stansberry commented on JBAS-7248: ---------------------------------------- I've added a simple hook to testsuite/imports/server-config.xml to make it easier to run the tests with JBoss Profiler from hudson (or the command line anywhere). Basically the value of property jboss.profiler.cmdline gets passed as a jvm argument to server classloader-leak and, for the clustered tests, classloader-leak-0. So: ./build.sh -Djboss.profiler.cmdline="-agentlib:jbossAgent" tests-all-classloader-leak or, perhaps, on a 64-bit jvm: ./build.sh -Djboss.profiler.cmdline="-agentlib:jbossAgent -d64" tests-all-classloader-leak The default value of the property is "". This just deals with passing the necessary command arguments to java for the servers. The other aspects of using JBP discussed at http://www.jboss.org/community/docs/DOC-9294 still need to be handled. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://jira.jboss.org/browse/JBAS-7248?page=com.atlassian.jira.plugin.sy... ] Shelly McGowan updated JBAS-7248: --------------------------------- Running the job with JBoss Profiler, the server encountered an OutOfMemoryError: java heap. Heap dump captured in the JBoss bin directory of the test Workspace. http://hudson.qa.jboss.com/hudson/view/JBoss%20AS/job/JBoss-AS-6.0.x-clas... -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://jira.jboss.org/browse/JBAS-7248?page=com.atlassian.jira.plugin.sy... ] Scott Marlow commented on JBAS-7248: ------------------------------------ Instead of using JBoss profiler to capture the memory heap, I wonder if it would be less pain, to use the jdk tools to generate a heapdump (perhaps via the Java -XX:+HeapDumpOnOutOfMemoryError option). Then, we can use the Eclipse Memory Analyzer to read the heap dump (should be easier to use the JBoss profiler at the moment). I would be happy to help analyze the heap dump with MAT, which is currently the best OSS tool for memory leaks. :) -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://jira.jboss.org/browse/JBAS-7248?page=com.atlassian.jira.plugin.sy... ] Scott Marlow commented on JBAS-7248: ------------------------------------ In order to have a 2 gig java heap, we will need to run on a 64bit Hudson slave machine that probably has a 3-4 gig OS memory size -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://jira.jboss.org/browse/JBAS-7248?page=com.atlassian.jira.plugin.sy... ] Shelly McGowan commented on JBAS-7248: -------------------------------------- The heap has been increased to 2048m and are tied to node dsp-mm1-rhel-64 (9GB RAM, 2 QUAD core, Xeon 1.6 Ghz). The test runs have not been completing - hanging during server shutdown. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira