Minor nitpick: Archive is called
apache-tomee-version.qualifier-edition.zip where the extracted
directory is called apache-tomee-edition-version.qualifier.
Cheers
Daniel
On Sat, Jan 7, 2012 at 11:59 PM, David Blevins david.blev...@gmail.com wrote:
Ok, binaries are ready for a vote! Have run
Had thought this was fixed, but Romain confirms it's still broken.
Best to just cut out the remote adapter if it doesn't work. We can try again
next release.
Had already re-rolled the release last night and ran the TCK, but will do that
again with this trimmed.
-David
On Jan 13, 2012, at
2 words about this issue:
1) it is still here and it is difficult to reproduce (sometimes it passes,
sometimes it fails)
2) there was a regression last week i reverted and now you can work on
windows too (on trunk)
- Romain
2012/1/15 David Blevins david.blev...@gmail.com
Had thought this was
I am now getting the following, is there a missing or modified default port
configuration somewhere?:
Running org.apache.openejb.arquillian.session.SessionScopeTest
[] START SERVER
OPENEJB_HOME =
D:\tmp\arquillian-apache-tomee\apache-tomee-plus-1.0.0-beta-2-SNAPSHOT
SYSTEM_INFO = Java 1.6.0_30;
The port should be 10080 by default.
can you check in :
openejb-trunk/arquillian-tomee/arquillian-tomee-remote/src/test/resources/arquillian.xml
- Romain
2012/1/13 AndyG andy.gumbre...@orprovision.com
I am now getting the following, is there a missing or modified default port
configuration
Still the same, something has changed since yesterday morning that is causing
the test tomee to try and bind to port 0 - Trying to figure out what.
--
View this message in context:
http://openejb.979440.n4.nabble.com/VOTE-OpenEJB-4-0-0-beta-2-TomEE-1-0-0-beta-2-tp4274468p4292230.html
Sent from
i think i did a if(true) instead of if(false) but was never in this case
under win. I'll fix it.
- Romain
2012/1/12 David Blevins david.blev...@gmail.com
On Jan 10, 2012, at 3:26 PM, Jonathan Gallimore wrote:
I added a system property: openejb.server.debug=true to
RemoteTomEEContainer,
On Jan 11, 2012, at 11:25 PM, David Blevins wrote:
On Jan 10, 2012, at 3:26 PM, Jonathan Gallimore wrote:
I added a system property: openejb.server.debug=true to
RemoteTomEEContainer, and I was then able to hook up a remote debugger to
the server. Stopping in
Had a similar fix in the pipes here, was just not quick enough on the
check-in. However, just for info, I have had to bump up a few build
environment options on Windows to get a full build to complete:
set MAVEN_OPTS=-Xmx768m -XX:MaxPermSize=256M -XX:ReservedCodeCacheSize=64m
-Xss4096k
The
he got me too ;)
- Romain
2012/1/12 AndyG andy.gumbre...@orprovision.com
Had a similar fix in the pipes here, was just not quick enough on the
check-in. However, just for info, I have had to bump up a few build
environment options on Windows to get a full build to complete:
set
On Jan 12, 2012, at 2:49 AM, AndyG wrote:
Had a similar fix in the pipes here, was just not quick enough on the
check-in. However, just for info, I have had to bump up a few build
environment options on Windows to get a full build to complete:
set MAVEN_OPTS=-Xmx768m -XX:MaxPermSize=256M
Hehe. Well, it'll be very late tonight before I'll get the chance to hack
again. I can port that fix over to the branch, or if someone wants to be beat
me to it I can spend that time rolling the new release binaries instead.
Either way, we're looking very good thanks to everyone who has
Maven OPTS set to (for WIn Sun SDK):
env={'MAVEN_OPTS': '-Xmx1024m -Xss4096k -XX:ReservedCodeCacheSize=64m
-XX:PermSize=64m -XX:MaxPermSize=256m -XX:+HeapDumpOnOutOfMemoryError
-XX:HeapDumpPath=C:\slave4\openejb-trunk-win\dumps'}
On Thu, Jan 12, 2012 at 9:04 PM, David Blevins
I would say most definitely related to the classloading issue. Only one
TestRunner is allowed per classloader.
The
D:/tmp/arquillian-apache-tomee/apache-tomee-plus-1.0.0-beta-2/temp/arquillian-junit-4211441189472437588.jar
is going to be on the application root classloader and all child
can you check in the classloader in
org.apache.tomee.catalina.TomEEWebappLoader.TomEEClassLoader please?
- Romain
2012/1/11 AndyG andy.gumbre...@orprovision.com
I would say most definitely related to the classloading issue. Only one
TestRunner is allowed per classloader.
The
I guess you mean the addIfNotExist method?
--
View this message in context:
http://openejb.979440.n4.nabble.com/VOTE-OpenEJB-4-0-0-beta-2-TomEE-1-0-0-beta-2-tp4274468p4285334.html
Sent from the OpenEJB Dev mailing list archive at Nabble.com.
was what i was thinking about but what is the issue with windows since it
works with unix OS(s)?
- Romain
2012/1/11 AndyG andy.gumbre...@orprovision.com
Having a play now, the URLs need to be normalized and compared more
explicitly against the temp URLs
--
View this message in context:
Different temp file handling?
On Wed, Jan 11, 2012 at 3:16 PM, Romain Manni-Bucau
rmannibu...@gmail.com wrote:
was what i was thinking about but what is the issue with windows since it
works with unix OS(s)?
- Romain
2012/1/11 AndyG andy.gumbre...@orprovision.com
Having a play now, the
Did you manage to know if it comes from
1) us
2) arquillian
3) windows itself
?
- Romain
2012/1/11 AndyG andy.gumbre...@orprovision.com
Results of a quick debug check:
On windows the arquillian-junit.jar is pushed to temp and renamed to:
[temp]/arquillian-junit-1163337363386056329.jar --
Could be
org.jboss.arquillian.container.jbossas.remote_6.ShrinkWrapUtil::toURL().
https://github.com/arquillian/arquillian-container-jbossas/blob/master/jbossas-remote-6/src/main/java/org/jboss/arquillian/container/jbossas/remote_6/ShrinkWrapUtil.java
Can you debug into that class?
Cheers
The whole temp file thing seems to have arisen from JDK inconsistency issues
related to the ClassLoaderUtil.clearSunJarFileFactoryCache method.
This method was for a period failing to actually release the file handle on
jar files due to some real spaghetti in the
On Jan 11, 2012, at 7:45 AM, AndyG wrote:
The whole temp file thing seems to have arisen from JDK inconsistency issues
related to the ClassLoaderUtil.clearSunJarFileFactoryCache method.
This method was for a period failing to actually release the file handle on
jar files due to some real
On Jan 10, 2012, at 10:52 AM, dsh wrote:
On the other hand I'd like to understand why building and sanity
testing a particular branch to be released on one of the windows
buildbot slaves isn't enough. Is it because each buildbot slave needs
to be manually configured for each new branch that
Ok if it's really just that the windows slave needs to be setup
accordingly then we should take that into consideration. That way we
would have CI feedback per checkin again. IIRC I even had an account
once on the windows buildbot slave and we did a complementary setup
for the IBM JDK which became
On Jan 11, 2012, at 12:06 PM, dsh wrote:
Ok if it's really just that the windows slave needs to be setup
accordingly then we should take that into consideration. That way we
would have CI feedback per checkin again. IIRC I even had an account
once on the windows buildbot slave and we did a
Done! (deleted the .m2 repo contents, the various SVN checkouts and
restarted the buildbot service)
Cheers
Daniel
On Wed, Jan 11, 2012 at 9:26 PM, David Blevins david.blev...@gmail.com wrote:
On Jan 11, 2012, at 12:06 PM, dsh wrote:
Ok if it's really just that the windows slave needs to be
The bb-2008 slave now seems to start compiling but it looks like we do
have tmp file issues:
http://ci.apache.org/builders/openejb-trunk-win-sunjdk/builds/1673/steps/compile/logs/stdio
Cheers
Daniel
On Wed, Jan 11, 2012 at 10:29 PM, David Blevins david.blev...@gmail.com wrote:
On Jan 11,
What I as well noticed is that some of the JUnit test XML files are
pretty long and thus can't be deleted under Windoze.
On Thu, Jan 12, 2012 at 12:13 AM, dsh daniel.hais...@googlemail.com wrote:
The bb-2008 slave now seems to start compiling but it looks like we do
have tmp file issues:
On Jan 10, 2012, at 3:26 PM, Jonathan Gallimore wrote:
I added a system property: openejb.server.debug=true to
RemoteTomEEContainer, and I was then able to hook up a remote debugger to
the server. Stopping in
org.jboss.arquillian.container.test.spi.util.ServiceLoader, it appear that
the
Don't know if that is due to my environment, but I have some arquilian
tests failing.
I notice as well 3 files containing SNAPSHOT instead of the beta-2.
./examples/cdi-alternative-and-stereotypes/README.md:Apache OpenEJB
4.0.0-beta-2-SNAPSHOTbuild: 20111030-07:54
On Jan 10, 2012, at 1:09 AM, Jean-Louis MONTEIRO wrote:
Don't know if that is due to my environment, but I have some arquilian
tests failing.
Can you paste the error? They pass in buildbot and on my machine, but that
doesn't mean there might not be an issue.
I notice as well 3 files
On Jan 9, 2012, at 2:14 PM, Jonathan Gallimore wrote:
Hi David,
Did you by any chance run the legal tool you wrote around the time of the
last release? I remember finding it really useful. I'm happy to give it a
run and post up the results if you give me a couple of pointers.
I'm just
How was this report generated? Can it be generated by everybody?
Cheers
Daniel
On Tue, Jan 10, 2012 at 12:02 PM, David Blevins david.blev...@gmail.com wrote:
On Jan 9, 2012, at 2:14 PM, Jonathan Gallimore wrote:
Hi David,
Did you by any chance run the legal tool you wrote around the time
Andy +0
Two test failures. Not dug in to the cause, so if anyone can see something
in the traces that may have something to do with my setup then let me know.
Using Win7 Pro 64bit
---
Test set:
More or less the same as Andy, but I'm on windows and my computer is still
very slow.
Dunno if that is the reason.
Jean-Louis
2012/1/10 AndyG andy.gumbre...@orprovision.com
Andy +0
Two test failures. Not dug in to the cause, so if anyone can see something
in the traces that may have
Stupid question. Why do we test locally? Can't we get VM images of the
most important operating systems for our testing that could be set
back to a canonical snapshot each time we intend to run a release
cycle?
I have such an environment at home but can't run it 7x24h cause of the
energy costs
to be sure it works for eveybody?
if we all use a vm it will work for eveybody and we can miss some important
issues.
- Romain
2012/1/10 dsh daniel.hais...@googlemail.com
Stupid question. Why do we test locally? Can't we get VM images of the
most important operating systems for our testing
IMO we should be testing specific environments and not local setups.
What's makes a local setup preferable over a canonical and clean setup
that can be reinitialized prior to each release engeneering cycle?
AFAIK the issues reported might be related to Windoze platforms. A
platform that seems to
if we can have win vm too +1
window$ is not so used excepted on dev computers :(
- Romain
2012/1/10 dsh daniel.hais...@googlemail.com
IMO we should be testing specific environments and not local setups.
What's makes a local setup preferable over a canonical and clean setup
that can be
Yes... this is the same issue I reported some days ago for the trunk.
I have the same issue in my windows and linux machines.
[]s,
Thiago.
On Tue, Jan 10, 2012 at 10:32 AM, Romain Manni-Bucau
rmannibu...@gmail.comwrote:
if we can have win vm too +1
window$ is not so used excepted on dev
Darn.
I might be able to dig in on Saturday or Sunday, but not likely sooner. If
anyone out there who has the problem is able to dig a little it'd be very
appreciated.
-David
On Jan 10, 2012, at 4:36 AM, AndyG wrote:
Andy +0
Two test failures. Not dug in to the cause, so if anyone
I think in theory, if each one of us has an MSDN account, we might be
able to have a windows VM that we could share amongst OEJB devs. I
think the only thing that's probably prohibited is making them
available in a public place such as people.a.o.
On the other hand I'd like to understand why
Going to take a look on my Windows machine here - I'll shout if I can see
what the problem is.
Jon
On Tue, Jan 10, 2012 at 6:45 PM, David Blevins david.blev...@gmail.comwrote:
Darn.
I might be able to dig in on Saturday or Sunday, but not likely sooner.
If anyone out there who has the
I added a system property: openejb.server.debug=true to
RemoteTomEEContainer, and I was then able to hook up a remote debugger to
the server. Stopping in
org.jboss.arquillian.container.test.spi.util.ServiceLoader, it appear that
the JUnitTestRunning is being picked up from 2 jars - one in the
+1
Cheers
Daniel
On Sat, Jan 7, 2012 at 11:59 PM, David Blevins david.blev...@gmail.com wrote:
Ok, binaries are ready for a vote! Have run the TCK on these and everything
looks good -- will post link to the tck@ list.
SVN Tag:
+1
On Jan 9, 2012 9:44 AM, dsh daniel.hais...@googlemail.com wrote:
+1
Cheers
Daniel
On Sat, Jan 7, 2012 at 11:59 PM, David Blevins david.blev...@gmail.com
wrote:
Ok, binaries are ready for a vote! Have run the TCK on these and
everything looks good -- will post link to the tck@ list.
Hi David,
Did you by any chance run the legal tool you wrote around the time of the
last release? I remember finding it really useful. I'm happy to give it a
run and post up the results if you give me a couple of pointers.
I'm just checking things over here, hope to post a vote soon.
Cheers
+1
- Romain
2012/1/8 Jean-Louis MONTEIRO jeano...@gmail.com
Will have a look on Monday morning.
Have a nice weekend.
Jean-Louis
2012/1/7 David Blevins david.blev...@gmail.com
Ok, binaries are ready for a vote! Have run the TCK on these and
everything looks good -- will post link to
Ok, binaries are ready for a vote! Have run the TCK on these and everything
looks good -- will post link to the tck@ list.
SVN Tag:
http://svn.apache.org/repos/asf/openejb/tags/openejb-4.0.0-beta-2/
Maven Repo:
https://repository.apache.org/content/repositories/orgapacheopenejb-029/
Will have a look on Monday morning.
Have a nice weekend.
Jean-Louis
2012/1/7 David Blevins david.blev...@gmail.com
Ok, binaries are ready for a vote! Have run the TCK on these and
everything looks good -- will post link to the tck@ list.
SVN Tag:
50 matches
Mail list logo