Well, things are a little bit better now.   

On my notebook, I ran "mvn clean install" in core 5 times:
   1)  All tests passed (that's a first)
   2-3) org.apache.yoko.TypesTest.testConst hung with the Shutdown in progress 
error
   4) All tests passed
   5) org.apache.yoko.TypesTest.testConst failed again

On pdceltix:  100% failure with 5 runs
   1-5) org.apache.yoko.processmanager.internal.ProcessAgentImpl.invokeStatic 
failed with 
java.io.FileNotFoundException: TestCodeSets.ref (No such file or directory)
[client:err] org.omg.CORBA.BAD_PARAM: bad scheme specific part: file error  
minor code: 0x4f4d0009  completed: No


One note: on the notebook, when the tests passed, they seem to have run in a 
different order than when they failed.   The kind of "smells" like  a test 
cleanup issue.   I added a "<forkMode>pertest</forkMode>" entry to the 
pom.xml to force a clean VM for each test and reran.   The tests then did 
pass on pdceltix twice (out of 2 attempts).   Still got the shutdown in 
progress error once (out of 3) on the notebook. 


Dan


On Wednesday August 23 2006 2:27 pm, Darren Middleman wrote:
> I've committed the fix for that failure.  Can you give it a try and let me
> know
> if it works for your distributions.
>
> Thanks,
> Darren
>
> On 8/23/06, Darren Middleman <[EMAIL PROTECTED]> wrote:
> > Hello again,
> >
> > I've managed to find the cause of the problem so I'm going to create a
> > JIRA
> > issue then submit the fix.  This should fix only the problem with that
> > one particular
> > test, however, since I am not able to duplicate the other that you are
> > seeing.
> >
> > Darren
> >
> > On 8/23/06, Darren Middleman <[EMAIL PROTECTED]> wrote:
> > > Hi,
> > >
> > > I'm seeing this too (Dual [EMAIL PROTECTED]).  The interesting thing is 
> > > that
> > > this test only
> > > fails as part of the Yoko test suite, never as part of the original
> > > IONA test where it
> > > came from so I'm wondering if it is something with the way the test is
> > > set up in Yoko.
> > > I'll try having a look into it.
> > >
> > > Darren
> > >
> > > On 8/23/06, Geir Magnusson Jr <[EMAIL PROTECTED]> wrote:
> > > > I thought I reported this a while ago...
> > > >
> > > > I saw the same thing on what appears to be a similar machine... a IBM
> > > > T43 and a T42...
> > > >
> > > > geir
> > > >
> > > > Daniel Kulp wrote:
> > > > > Matteo,
> > > > >
> > > > > I think it's more than intermittent, at least for me.    I just
> > > >
> > > > tried "mvn
> > > >
> > > > > clean install" in core 10 times on my machine (Intel(R) Pentium(R)
> > > > > M processor 2.13GHz) and all 10 times it hung on one of the two
> > > > > tests.
> > > > >
> > > > > I also tried on pdceltix (internal iona machine) which is a dual
> > > >
> > > > Intel(R)
> > > >
> > > > > Xeon(TM) CPU 3.00GHz.    It failed 5 out 5 times, but as you said,
> > > >
> > > > on
> > > >
> > > > > different tests this time.
> > > > >
> > > > > Basically, I CANNOT build the core without turning off the tests.
> > > >
> > > > In my
> > > >
> > > > > opinion, the tests should be disabled until they can be
> > > >
> > > > investigated.   As it
> > > >
> > > > > stands now, I'd have to say -1 (non-binding) to the release
> > > >
> > > > candidate just
> > > >
> > > > > because I cannot build it.
> > > > >
> > > > >
> > > > > Dan
> > > > >
> > > > > On Wednesday August 23 2006 11:10 am, Vescovi, Matteo wrote:
> > > > >> Hi,
> > > > >> I have seen this issue before.
> > > > >>
> > > > >> Here is a JIRA issue:
> > > > >> https://issues.apache.org/jira/browse/YOKO-129
> > > > >>
> > > > >> The problem is not always reproducible and happens inconsistently.
> > > > >> Most test runs complete successfully for me. The test runs that
> > > >
> > > > fail,
> > > >
> > > > >> fail on different tests.
> > > > >>
> > > > >> I tested on a slower uniprocessor machine, and the problem appears
> > > >
> > > > more
> > > >
> > > > >> frequently there.
> > > > >>
> > > > >> Everything seems to suggest a threading issue.
> > > > >>
> > > > >> - Matteo
> > > > >>
> > > > >> Daniel Kulp wrote:
> > > > >>> On my Linux box with JDK 1.5.0_08, if I run "mvn install" top
> > > >
> > > > level (fresh
> > > >
> > > > >>> checkout of the code from SVN), the tests hang in the core
> > > > >>> module.
> > > > >>>
> > > > >>> Output is below.
> > > > >>>
> > > > >>> Also, while building, I noticed the yoko-spec-corba stuff is
> > > > >>> being
> > > >
> > > > built
> > > >
> > > > >>> with "source/target=1.5" instead of 1.4.   I believe that should
> > > >
> > > > also be
> > > >
> > > > >>> changed.

-- 
J. Daniel Kulp
Principal Engineer
IONA
P: 781-902-8727    C: 508-380-7194   F:781-902-8001
[EMAIL PROTECTED]

Reply via email to