9:47 a.m.
On 4 Jan 2012, at 4:40 PM, Nicolas Helleringer wrote:
Yes
They are +very+ local hardware/environnement sensitive
It definitely seems like it.
--Hardy
10:11 a.m.
New subject: Hibernate search master build issues
Yes since that pull, which Emmanuel merged (so I guess it's working
fine for him too) it's consistently broken for me, and for our Jenkins
instances too.
I spent and hour on it yesterday but couldn't understand it. The
failure I have is HornetQ not being able to boot in AS7, then our test
fails as it depends on JMS.; It fails consistently when using the
build.sh script which forks my repository before starting the tests,
but I noticed that it occasionally doesn't fail if I run "mvn"
directly.
Running mvn directly means running the tests on an ext4 partition;
when I use the build script it actually clones to a ramfs mount;
looking at HornetQ's JIRA for issues I understood that its native
component is indeed partition-sensitive so I was about to open an
issue there... but then it started failing me on ext4 as well so I'm
back to square 1: clueless.
Still since it's AS7 failing to boot a service way before we even
start deploying our test WAR, I wouldn't blame our test... just weird
that this is broken after your test changes. Now if only you hadn't
reformatted all whitespace I would be glad to compare the files :) I
will resume the inspection later.
Cheers,
Sanne
On 4 September 2012 16:47, Hardy Ferentschik <hardy(a)hibernate.org> wrote:
On 4 Jan 2012, at 4:40 PM, Nicolas Helleringer wrote:
> Yes
>
> They are +very+ local hardware/environnement sensitive
It definitely seems like it.
--Hardy
_______________________________________________
hibernate-dev mailing list
hibernate-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
10:32 a.m.
New subject: Hibernate search master build issues
On 4 Jan 2012, at 5:11 PM, Sanne Grinovero wrote:
I spent and hour on it yesterday but couldn't understand it. The
failure I have is HornetQ not being able to boot in AS7, then our test
fails as it depends on JMS.; It fails consistently when using the
build.sh script which forks my repository before starting the tests,
but I noticed that it occasionally doesn't fail if I run "mvn"
directly.
Running mvn directly means running the tests on an ext4 partition;
when I use the build script it actually clones to a ramfs mount;
looking at HornetQ's JIRA for issues I understood that its native
component is indeed partition-sensitive so I was about to open an
issue there... but then it started failing me on ext4 as well so I'm
back to square 1: clueless.
Still since it's AS7 failing to boot a service way before we even
start deploying our test WAR, I wouldn't blame our test... just weird
that this is broken after your test changes. Now if only you hadn't
reformatted all whitespace I would be glad to compare the files :) I
will resume the inspection later.
My initial guess for the new failures wold have been my change to
RegistrationConfiguration#createTempDir. I thought maybe the temp
directory does not get deleted properly.
But then again, I have not seen the error yet and it does not seem related
to HornetQ.
--Hardy
4494
days inactive
4494
days old
2 comments
2 participants
participants (2)
-
Hardy Ferentschik -
Sanne Grinovero