You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@deltaspike.apache.org by Andrew Bayer <ab...@apache.org> on 2015/12/30 16:57:14 UTC
DeltaSpike and Jenkins - please stop running excessive builds
I discovered today while checking on load on builds.apache.org that not
only does DeltaSpike have a huge number of jobs that poll and run whenever
there's a change in Git - all those jobs are then kicked off *again* by
https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked off by
the same polling! The end result is running 50 or so jobs *twice*, taking
half an hour or so each, for every single change that goes into DeltaSpike.
Please, please, please fix this to only trigger your jobs once each, not
twice, and ideally, it'd be great if you could rationalize your jobs and
cut down the number of jobs, so as not to hog basically all the available
slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
A.
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by Gerhard Petracek <gp...@apache.org>.
@john:
harald wrote about the inconsistent jenkins-setup some days ago.
(it just takes some time to align the execution, because we have/need that
many jenkins-jobs.)
regards,
gerhard
2015-12-30 17:03 GMT+01:00 John D. Ament <jo...@apache.org>:
> Where was it discussed?
>
> On Wed, Dec 30, 2015 at 11:03 AM Gerhard Petracek <gp...@apache.org>
> wrote:
>
> > hi andrew,
> >
> > we have discussed it recently and we are going to change it soon.
> >
> > regards,
> > gerhard
> >
> >
> >
> > 2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
> >
> > > I discovered today while checking on load on builds.apache.org that
> not
> > > only does DeltaSpike have a huge number of jobs that poll and run
> > whenever
> > > there's a change in Git - all those jobs are then kicked off *again* by
> > > https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked
> off
> > by
> > > the same polling! The end result is running 50 or so jobs *twice*,
> taking
> > > half an hour or so each, for every single change that goes into
> > DeltaSpike.
> > > Please, please, please fix this to only trigger your jobs once each,
> not
> > > twice, and ideally, it'd be great if you could rationalize your jobs
> and
> > > cut down the number of jobs, so as not to hog basically all the
> available
> > > slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
> > >
> > > A.
> > >
> >
>
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by "John D. Ament" <jo...@apache.org>.
Where was it discussed?
On Wed, Dec 30, 2015 at 11:03 AM Gerhard Petracek <gp...@apache.org>
wrote:
> hi andrew,
>
> we have discussed it recently and we are going to change it soon.
>
> regards,
> gerhard
>
>
>
> 2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
>
> > I discovered today while checking on load on builds.apache.org that not
> > only does DeltaSpike have a huge number of jobs that poll and run
> whenever
> > there's a change in Git - all those jobs are then kicked off *again* by
> > https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked off
> by
> > the same polling! The end result is running 50 or so jobs *twice*, taking
> > half an hour or so each, for every single change that goes into
> DeltaSpike.
> > Please, please, please fix this to only trigger your jobs once each, not
> > twice, and ideally, it'd be great if you could rationalize your jobs and
> > cut down the number of jobs, so as not to hog basically all the available
> > slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
> >
> > A.
> >
>
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by Andrew Bayer <an...@gmail.com>.
Thanks - much appreciated.
A.
On Wed, Dec 30, 2015 at 1:52 PM, Gerhard Petracek <gp...@apache.org>
wrote:
> @andrew:
> it should be fine now.
>
> regards,
> gerhard
>
>
>
> 2015-12-30 17:10 GMT+01:00 Gerhard Petracek <gp...@apache.org>:
>
>> @andrew:
>> i'll do it within the next hours.
>>
>> regards,
>> gerhard
>>
>>
>>
>> 2015-12-30 17:05 GMT+01:00 Andrew Bayer <an...@gmail.com>:
>>
>>> Please change at least the triggering ASAP, or I'll disable all the
>>> DeltaSpike jobs to free up capacity for other projects until it's fixed.
>>> Sorry to be harsh, but this is a colossal use of limited build resources.
>>>
>>> A.
>>>
>>> On Wed, Dec 30, 2015 at 11:02 AM, Gerhard Petracek <gpetracek@apache.org
>>> > wrote:
>>>
>>>> hi andrew,
>>>>
>>>> we have discussed it recently and we are going to change it soon.
>>>>
>>>> regards,
>>>> gerhard
>>>>
>>>>
>>>>
>>>> 2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
>>>>
>>>>> I discovered today while checking on load on builds.apache.org that
>>>>> not
>>>>> only does DeltaSpike have a huge number of jobs that poll and run
>>>>> whenever
>>>>> there's a change in Git - all those jobs are then kicked off *again* by
>>>>> https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked
>>>>> off by
>>>>> the same polling! The end result is running 50 or so jobs *twice*,
>>>>> taking
>>>>> half an hour or so each, for every single change that goes into
>>>>> DeltaSpike.
>>>>> Please, please, please fix this to only trigger your jobs once each,
>>>>> not
>>>>> twice, and ideally, it'd be great if you could rationalize your jobs
>>>>> and
>>>>> cut down the number of jobs, so as not to hog basically all the
>>>>> available
>>>>> slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
>>>>>
>>>>> A.
>>>>>
>>>>
>>>>
>>>
>>
>
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by Gerhard Petracek <gp...@apache.org>.
@andrew:
it should be fine now.
regards,
gerhard
2015-12-30 17:10 GMT+01:00 Gerhard Petracek <gp...@apache.org>:
> @andrew:
> i'll do it within the next hours.
>
> regards,
> gerhard
>
>
>
> 2015-12-30 17:05 GMT+01:00 Andrew Bayer <an...@gmail.com>:
>
>> Please change at least the triggering ASAP, or I'll disable all the
>> DeltaSpike jobs to free up capacity for other projects until it's fixed.
>> Sorry to be harsh, but this is a colossal use of limited build resources.
>>
>> A.
>>
>> On Wed, Dec 30, 2015 at 11:02 AM, Gerhard Petracek <gp...@apache.org>
>> wrote:
>>
>>> hi andrew,
>>>
>>> we have discussed it recently and we are going to change it soon.
>>>
>>> regards,
>>> gerhard
>>>
>>>
>>>
>>> 2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
>>>
>>>> I discovered today while checking on load on builds.apache.org that not
>>>> only does DeltaSpike have a huge number of jobs that poll and run
>>>> whenever
>>>> there's a change in Git - all those jobs are then kicked off *again* by
>>>> https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked
>>>> off by
>>>> the same polling! The end result is running 50 or so jobs *twice*,
>>>> taking
>>>> half an hour or so each, for every single change that goes into
>>>> DeltaSpike.
>>>> Please, please, please fix this to only trigger your jobs once each, not
>>>> twice, and ideally, it'd be great if you could rationalize your jobs and
>>>> cut down the number of jobs, so as not to hog basically all the
>>>> available
>>>> slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
>>>>
>>>> A.
>>>>
>>>
>>>
>>
>
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by Gerhard Petracek <gp...@apache.org>.
@andrew:
i'll do it within the next hours.
regards,
gerhard
2015-12-30 17:05 GMT+01:00 Andrew Bayer <an...@gmail.com>:
> Please change at least the triggering ASAP, or I'll disable all the
> DeltaSpike jobs to free up capacity for other projects until it's fixed.
> Sorry to be harsh, but this is a colossal use of limited build resources.
>
> A.
>
> On Wed, Dec 30, 2015 at 11:02 AM, Gerhard Petracek <gp...@apache.org>
> wrote:
>
>> hi andrew,
>>
>> we have discussed it recently and we are going to change it soon.
>>
>> regards,
>> gerhard
>>
>>
>>
>> 2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
>>
>>> I discovered today while checking on load on builds.apache.org that not
>>> only does DeltaSpike have a huge number of jobs that poll and run
>>> whenever
>>> there's a change in Git - all those jobs are then kicked off *again* by
>>> https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked off
>>> by
>>> the same polling! The end result is running 50 or so jobs *twice*, taking
>>> half an hour or so each, for every single change that goes into
>>> DeltaSpike.
>>> Please, please, please fix this to only trigger your jobs once each, not
>>> twice, and ideally, it'd be great if you could rationalize your jobs and
>>> cut down the number of jobs, so as not to hog basically all the available
>>> slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
>>>
>>> A.
>>>
>>
>>
>
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by Andrew Bayer <an...@gmail.com>.
Please change at least the triggering ASAP, or I'll disable all the
DeltaSpike jobs to free up capacity for other projects until it's fixed.
Sorry to be harsh, but this is a colossal use of limited build resources.
A.
On Wed, Dec 30, 2015 at 11:02 AM, Gerhard Petracek <gp...@apache.org>
wrote:
> hi andrew,
>
> we have discussed it recently and we are going to change it soon.
>
> regards,
> gerhard
>
>
>
> 2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
>
>> I discovered today while checking on load on builds.apache.org that not
>> only does DeltaSpike have a huge number of jobs that poll and run whenever
>> there's a change in Git - all those jobs are then kicked off *again* by
>> https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked off
>> by
>> the same polling! The end result is running 50 or so jobs *twice*, taking
>> half an hour or so each, for every single change that goes into
>> DeltaSpike.
>> Please, please, please fix this to only trigger your jobs once each, not
>> twice, and ideally, it'd be great if you could rationalize your jobs and
>> cut down the number of jobs, so as not to hog basically all the available
>> slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
>>
>> A.
>>
>
>
Re: DeltaSpike and Jenkins - please stop running excessive builds
Posted by Gerhard Petracek <gp...@apache.org>.
hi andrew,
we have discussed it recently and we are going to change it soon.
regards,
gerhard
2015-12-30 16:57 GMT+01:00 Andrew Bayer <ab...@apache.org>:
> I discovered today while checking on load on builds.apache.org that not
> only does DeltaSpike have a huge number of jobs that poll and run whenever
> there's a change in Git - all those jobs are then kicked off *again* by
> https://builds.apache.org/job/DeltaSpike%20Deploy/, which is kicked off by
> the same polling! The end result is running 50 or so jobs *twice*, taking
> half an hour or so each, for every single change that goes into DeltaSpike.
> Please, please, please fix this to only trigger your jobs once each, not
> twice, and ideally, it'd be great if you could rationalize your jobs and
> cut down the number of jobs, so as not to hog basically all the available
> slots for non-Hadoop-related jobs on builds.apache.org. Thanks.
>
> A.
>