You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@openjpa.apache.org by Mark Struberg <st...@yahoo.de> on 2015/06/03 19:30:09 UTC
freaking stochastic test
Testergebnis (1 fehlgeschlagener Test / +1)
• org.apache.openjpa.persistence.datacache.TestDataCacheBehavesIdentical.testDeleteIsNotDetectedOnCleanRefreshWithoutLockWithDataCache
This test fails randomly. On Windows more often, on OSX and Linux not that often but still.
The CI box seems stochastic alltogether. We had 150 failing tests and now only 1.
I gonna exclude this very test.
LieGrue,
strub
Re: freaking stochastic test
Posted by Mark Struberg <st...@yahoo.de>.
Yes, I had green builds on Windows as well. But it failes pretty often. Otoh I have no clue at what to look. This is in an area I’ve never looked at.
LieGrue,
strub
> Am 04.06.2015 um 20:04 schrieb Rick Curtis <cu...@gmail.com>:
>
> FWIW, I had a green build this morning on a windows box. I'm okay keeping
> an eye on it for another day to see what happens.
>
> On Wed, Jun 3, 2015 at 2:29 PM, Mark Struberg <st...@yahoo.de> wrote:
>
>> Good news, the build is green again!
>>
>> How should we proceed? I'd rather let it run for a few day and look
>> whether there are more stochastic failures. Might be because of a too low
>> ulimit for file handles? The passing tests was on ubuntu3, the failing ones
>> on ubuntu-1404. Not sure how much impact that has.
>>
>> In any case this DataCache test should get enabled again after a while.
>> And fixed of course...
>>
>> LieGrue,
>> strub
>>
>>
>>> Am 03.06.2015 um 19:30 schrieb Mark Struberg <st...@yahoo.de>:
>>>
>>> Testergebnis (1 fehlgeschlagener Test / +1)
>>> *
>> org.apache.openjpa.persistence.datacache.TestDataCacheBehavesIdentical.testDeleteIsNotDetectedOnCleanRefreshWithoutLockWithDataCache
>>>
>>> This test fails randomly. On Windows more often, on OSX and Linux not
>> that often but still.
>>>
>>> The CI box seems stochastic alltogether. We had 150 failing tests and
>> now only 1.
>>>
>>> I gonna exclude this very test.
>>>
>>> LieGrue,
>>> strub
>>
>>
>
>
> --
> *Rick Curtis*
Re: freaking stochastic test
Posted by Rick Curtis <cu...@gmail.com>.
FWIW, I had a green build this morning on a windows box. I'm okay keeping
an eye on it for another day to see what happens.
On Wed, Jun 3, 2015 at 2:29 PM, Mark Struberg <st...@yahoo.de> wrote:
> Good news, the build is green again!
>
> How should we proceed? I'd rather let it run for a few day and look
> whether there are more stochastic failures. Might be because of a too low
> ulimit for file handles? The passing tests was on ubuntu3, the failing ones
> on ubuntu-1404. Not sure how much impact that has.
>
> In any case this DataCache test should get enabled again after a while.
> And fixed of course...
>
> LieGrue,
> strub
>
>
> > Am 03.06.2015 um 19:30 schrieb Mark Struberg <st...@yahoo.de>:
> >
> > Testergebnis (1 fehlgeschlagener Test / +1)
> > *
> org.apache.openjpa.persistence.datacache.TestDataCacheBehavesIdentical.testDeleteIsNotDetectedOnCleanRefreshWithoutLockWithDataCache
> >
> > This test fails randomly. On Windows more often, on OSX and Linux not
> that often but still.
> >
> > The CI box seems stochastic alltogether. We had 150 failing tests and
> now only 1.
> >
> > I gonna exclude this very test.
> >
> > LieGrue,
> > strub
>
>
--
*Rick Curtis*
Re: freaking stochastic test
Posted by Mark Struberg <st...@yahoo.de>.
Good news, the build is green again!
How should we proceed? I’d rather let it run for a few day and look whether there are more stochastic failures. Might be because of a too low ulimit for file handles? The passing tests was on ubuntu3, the failing ones on ubuntu-1404. Not sure how much impact that has.
In any case this DataCache test should get enabled again after a while. And fixed of course…
LieGrue,
strub
> Am 03.06.2015 um 19:30 schrieb Mark Struberg <st...@yahoo.de>:
>
> Testergebnis (1 fehlgeschlagener Test / +1)
> • org.apache.openjpa.persistence.datacache.TestDataCacheBehavesIdentical.testDeleteIsNotDetectedOnCleanRefreshWithoutLockWithDataCache
>
> This test fails randomly. On Windows more often, on OSX and Linux not that often but still.
>
> The CI box seems stochastic alltogether. We had 150 failing tests and now only 1.
>
> I gonna exclude this very test.
>
> LieGrue,
> strub