This is because another Hudson job
(
http://gf-hudson.sfbay/hudson/job/gf-trunk-main-publisher) is trying to
publish the SNAPSHOT artifacts to the internal remote maven repository
and it's referencing the same local maven repo. If the continuous job
is triggered while the main-publisher job is still running, you'll see
the error that continuous job is unable to clean up the local repository.
Snjezana Sevo-Zenzerovic wrote:
> Speaking of Hudson glitches, any idea why is continuous job
> complaining for second time that it is unable to clean up some of
> local repository directories?
>
> Jane Young wrote:
>> There are Hudson glitches that may happen but if you take a look at
>> the console or where the build/test is failing you have a hunch if
>> it's related to your commit. So please run QL test on both profiles
>> (web and glassfish) before committing and monitor the Hudson job
>> (http://hudson.glassfish.org/job/gf-trunk-build-continuous/). Also
>> when your commit breaks the build or QL, you'll be notified via
>> e-mail. Some people have been ignoring e-mails from "RE Hudson".
>>
>> There were multiple commits that contributed to the QL failures this
>> morning. Thank you Jerome for taking the time to fix this.
>> Jane
>>
>>
>>
>> Jerome Dochez wrote:
>>> it's blue now, I think previous build was just a fluke.
>>> so you have nothing else to do
>>>
>>> and yes, I did talk about running the QL, not random hudson/build
>>> glitches that may or may not happen.
>>>
>>> jerome
>>>
>>> On Oct 21, 2009, at 2:48 PM, Ken Cavanaugh wrote:
>>>
>>>>
>>>> On Oct 21, 2009, at 2:44 PM, Jerome Dochez wrote:
>>>>
>>>>>
>>>>> On Oct 21, 2009, at 2:33 PM, Tim Quinn wrote:
>>>>>
>>>>>> Hi, Jerome.
>>>>>>
>>>>>> I know the feeling. Not that I've had to clean up anyone else's
>>>>>> QL breakages, but similarly I feel like over the years I've spent
>>>>>> months of time diagnosing some bug as someone else's problem.
>>>>>>
>>>>>> Anyway, I'm probably going to check in several changes later this
>>>>>> afternoon. I'll have run QL beforehand, as I always do, and I'll
>>>>>> monitor the hudson build as I always do. If I break it, I'll fix
>>>>>> it.
>>>>>>
>>>>>> But I'll be out of the office after the close of business today,
>>>>>> returning Monday morning. So if by some remote chance :-) I
>>>>>> broke the build, I will fix it but you and everyone else would
>>>>>> have a long wait for me to fix everyone else's breakages for the
>>>>>> next three days!
>>>>> don't worry, you next task will await you when you return in case
>>>>> you break it :)
>>>>>
>>>>> and you can always hope that no one else breaks it in the next
>>>>> three days too !
>>>>
>>>> Take a look at the job and my previous email. It's broken in a way
>>>> seemingly
>>>> unrelated to my commit, and I can't fix it.
>>>>
>>>> Now what?
>>>>
>>>> Ken.
>>>
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: dev-help_at_glassfish.dev.java.net
>