[hibernate-dev] Infinispan tests

[Sanne Grinovero]

On 8 March 2012 13:02, Steve Ebersole <steve at hibernate.org> wrote:
> I saw a file show up in one of them during the runs.  It was deleted
> afterwards.  As to whether or not they show up ever again if I manually
> delete them... dunno.  But I'd also like to know the who/why/whats
> about them as well...

Those are recovery logs from the JBoss TransactionManager, which are
flushed to disk to be able to recover prepared but uncommitted
transactions in case of a crash in between.

In Hibernate OGM and Search we had to add a configuration file in the
test resource root to disable the on-disk log:
https://github.com/hibernate/hibernate-search/blob/master/hibernate-search-integrationtest/src/test/resources/jbossts-properties.xml

I was going to add the same file in the hibernate-infinispan module
but I couldn't reproduce it.

>
> 11 minutes is less then what I saw.  This is not a "quick turnaround"
> and to be honest its the sole reason I do not full tests (as in running
> tests across all modules) before every checkin.  We need to think of
> ways to minimize this time.  As I understand it, the time issue it is a
> lot of wait statements related to setting up clusters for clustered
> testing support.  What if we moved clustering-related tests somewhere
> where they arent run every test run?  Maybe to intgTest?  Or to another
> git repo?  As I understand it the waits were introduced to try to
> minimize the random test failures.
>
> Sanne, yes its about moving to a non-Final release of a dependency
> during a point release.

Right; it would have been enough to move to Infinispan 5.1.2.FINAL
then as it was released before ORM, but I'd check that with Galder
first.

-- Sanne

>
>
> On Thu 08 Mar 2012 04:54:57 AM CST, Hardy Ferentschik wrote:
>>
>> On Mar 8, 2012, at 11:48 AM, Sanne Grinovero wrote:
>>
>>> I found the same files in my workspace as left over from and older
>>> build, but deleted them and tried new builds several times, they don't
>>> seem to be re-created anymore?
>>> Could anyone confirm (after cleaning them up) that it's still a
>>> problem, as I can't..
>>
>> i deleted them as well and they don't seem to get recreated.
>>
>> --Hardy
>> _______________________________________________
>> hibernate-dev mailing list
>> hibernate-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/hibernate-dev
>
> --
> steve at hibernate.org
> http://hibernate.org
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev