Bugs item #636227, was opened at 2002-11-10 16:05 You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=376685&aid=636227&group_id=22866
Category: JBossTest Group: CVS HEAD Status: Open Resolution: None Priority: 5 Submitted By: Chris Kimpton (kimptoc) Assigned to: Adrian Brock (ejort) Summary: testsuite generally fails to complete Initial Comment: Hi, I have run the full testsuite (testsuite/build.xml, target test) several times but it rarely completes. The machine is doing nothing else - 2 x 1.3GHZ cpu, 1GB ram - left run for 15 hours plus. Redhat 7.2, sun jdk1.3.1_06 Tried using kill -SIGHUP, but it seemed to kill the JVM rathing giving a thread dump. Regards, Chris ---------------------------------------------------------------------- >Comment By: Chris Kimpton (kimptoc) Date: 2002-11-11 14:04 Message: Logged In: YES user_id=39204 Hi, Yes - it does seem a little random. The full details are on this website: http://jboss.kimptoc.net/jboss-all [don't worry - it is jboss-head not all...] I was running jboss direct from CVS - no modifications, extra jars or config. Test log is here: http://jboss.kimptoc.net/jboss-all/build/cronjob_test.log The specific report you mentioned is here: http://jboss.kimptoc.net/jboss- all/testsuite/output/reports/TEST- org.jboss.test.naming.test.ExternalContextUnitTestCase.xml I can run it again later and capture the pstree output. Although I am pretty sure I see the ant java process in ps - as I have to kill them all off manually when this happens... Chris ---------------------------------------------------------------------- Comment By: Adrian Brock (ejort) Date: 2002-11-11 12:09 Message: Logged In: YES user_id=9459 That's a bit random. This test does nothing unless you have external contexts configured. What do you get in output/reports/TEST- org.jboss.test.naming.test.ExternalContextUnitTestCase.xml I'm beginning to suspect a bug in ant/junit especially since you are seeing a hang rather than timeout. What output do you get from pstree, can you see ant-java forking java? Do you have any jars in jre/lib/ext? Regards, Adrian ---------------------------------------------------------------------- Comment By: Chris Kimpton (kimptoc) Date: 2002-11-11 08:35 Message: Logged In: YES user_id=39204 2002-11-10 20:28:29,005 DEBUG [org.jboss.test.naming.test.ExternalContextUnitTestCase] JBossTestServices(), className=org.jboss.test.naming.test.ExternalContextUnitTestCase 2002- 11-10 20:28:29,368 DEBUG [org.jboss.test.naming.test.ExternalContextUnitTestCase] JBossTestServices(), className=org.jboss.test.naming.test.ExternalContextUnitTestCase 2002- 11-10 20:28:29,803 DEBUG [org.jboss.test.naming.test.ExternalContextUnitTestCase] No ExternalContext names exist ---------------------------------------------------------------------- Comment By: Adrian Brock (ejort) Date: 2002-11-10 23:05 Message: Logged In: YES user_id=9459 Hi Chris, Can you attach testsuite/output/log/test.log from the hanging test. Regards, Adrian ---------------------------------------------------------------------- Comment By: Chris Kimpton (kimptoc) Date: 2002-11-10 22:31 Message: Logged In: YES user_id=39204 Hi, I just tried running the tests-unit target on a freshly checked out set of code and still it hangs... I tried SIGQUIT - on several main java processes - but still did not see any thread dump in the server.log - or does it go to the console log? Thanks, Chris ---------------------------------------------------------------------- Comment By: Scott M Stark (starksm) Date: 2002-11-10 17:05 Message: Logged In: YES user_id=175228 SIGQUIT is the correct signal to use to generate a thread dump. ---------------------------------------------------------------------- Comment By: Adrian Brock (ejort) Date: 2002-11-10 16:42 Message: Logged In: YES user_id=9459 Hi Chris, Can you try it again? I saw a deadlock with OIL yesterday, it was actually a deadlock across two VMs during the connection close. I saw it as a timeout rather than a permenant hang though. I also found a problem where the dead letter queue wasn't being used because of incorrect xml parsing. This led to infinite redelivery of messages to MDBs. I've committed these fixes now. Regards, Adrian ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=376685&aid=636227&group_id=22866 ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
