You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@bigtop.apache.org by Evans Ye <ev...@apache.org> on 2016/01/07 18:55:22 UTC
Re: CI Bigtop-trunk job
I also noticed that.
What I've done was, I notice that the nexus container was gone on slave06,
and all the jobs start failing because there's no nexus exist, for example:
http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages/BUILD_ENVIRONMENTS=ubuntu-14.04,COMPONENTS=datafu,label=docker-slave/114/console
Since there're too many exited containers on slave06 as well, I therefore
destroy all the containers including running and exited on slave06.
I don't know why nexus container was gone, but Bigtop-trunk-packages
<http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages/>downstream
job looks strange to me:
Downstream Projects
- [image: Failed]Bigtop-trunk-packages-nexus-stop
<http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages-nexus-stop/>
If "nexus stop" is followed by packages, then no one is going to bring
nexus back, am I right?
2016-01-08 1:29 GMT+08:00 Olaf Flebbe <of...@oflebbe.de>:
> Hi,
>
> I am trying to understand what's going on with our CI.
>
> The Bigtop-trunk jobs all died on 7th January at ~ 14.30 UTC with an
> mysterious error message:
>
> Error response from daemon: Unknown device
> 311fdb8316738c37ea35a7c32a12e630dc0f24f05d804442a958e67db292276e
> (With varying hash nums)
>
> Did anyone do something weird on docker-slave-06 ?
>
> Olaf
>
>
Re: CI Bigtop-trunk job
Posted by Olaf Flebbe <of...@oflebbe.de>.
Hi,
starting and stopping the nexus does not really work as expected (Since I need to start nexus on _all_ slaves, otherwise one cannot use docker container networking).
So I wanted to try wether the upgrading the docker software fixes the problems on Bigtop-trunk (The all in one compile). Unfortunately that get killed.
I propose to remove the nexus workaround and look wether Bigtop-trunk and Bigtop-trunk-packages do work now without nexus.
Olaf
> Am 07.01.2016 um 19:23 schrieb Konstantin Boudnik <co...@apache.org>:
>
> On Fri, Jan 08, 2016 at 01:55AM, Evans Ye wrote:
>> I also noticed that.
>> What I've done was, I notice that the nexus container was gone on slave06,
>> and all the jobs start failing because there's no nexus exist, for example:
>>
>> http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages/BUILD_ENVIRONMENTS=ubuntu-14.04,COMPONENTS=datafu,label=docker-slave/114/console
>>
>> Since there're too many exited containers on slave06 as well, I therefore
>> destroy all the containers including running and exited on slave06.
>>
>> I don't know why nexus container was gone, but Bigtop-trunk-packages
>> <http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages/>downstream
>> job looks strange to me:
>>
>> Downstream Projects
>>
>> - [image: Failed]Bigtop-trunk-packages-nexus-stop
>> <http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages-nexus-stop/>
>>
>> If "nexus stop" is followed by packages, then no one is going to bring
>> nexus back, am I right?
>
> That's the oddity I have noticed as well and then it slipped through cracks. I
> think this should be done as a "After build step", when the whole matrix is
> finished.
>
> Cos
>
>> 2016-01-08 1:29 GMT+08:00 Olaf Flebbe <of...@oflebbe.de>:
>>
>>> Hi,
>>>
>>> I am trying to understand what's going on with our CI.
>>>
>>> The Bigtop-trunk jobs all died on 7th January at ~ 14.30 UTC with an
>>> mysterious error message:
>>>
>>> Error response from daemon: Unknown device
>>> 311fdb8316738c37ea35a7c32a12e630dc0f24f05d804442a958e67db292276e
>>> (With varying hash nums)
>>>
>>> Did anyone do something weird on docker-slave-06 ?
>>>
>>> Olaf
>>>
>>>
Re: CI Bigtop-trunk job
Posted by Konstantin Boudnik <co...@apache.org>.
On Fri, Jan 08, 2016 at 01:55AM, Evans Ye wrote:
> I also noticed that.
> What I've done was, I notice that the nexus container was gone on slave06,
> and all the jobs start failing because there's no nexus exist, for example:
>
> http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages/BUILD_ENVIRONMENTS=ubuntu-14.04,COMPONENTS=datafu,label=docker-slave/114/console
>
> Since there're too many exited containers on slave06 as well, I therefore
> destroy all the containers including running and exited on slave06.
>
> I don't know why nexus container was gone, but Bigtop-trunk-packages
> <http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages/>downstream
> job looks strange to me:
>
> Downstream Projects
>
> - [image: Failed]Bigtop-trunk-packages-nexus-stop
> <http://ci.bigtop.apache.org/view/Packages/job/Bigtop-trunk-packages-nexus-stop/>
>
> If "nexus stop" is followed by packages, then no one is going to bring
> nexus back, am I right?
That's the oddity I have noticed as well and then it slipped through cracks. I
think this should be done as a "After build step", when the whole matrix is
finished.
Cos
> 2016-01-08 1:29 GMT+08:00 Olaf Flebbe <of...@oflebbe.de>:
>
> > Hi,
> >
> > I am trying to understand what's going on with our CI.
> >
> > The Bigtop-trunk jobs all died on 7th January at ~ 14.30 UTC with an
> > mysterious error message:
> >
> > Error response from daemon: Unknown device
> > 311fdb8316738c37ea35a7c32a12e630dc0f24f05d804442a958e67db292276e
> > (With varying hash nums)
> >
> > Did anyone do something weird on docker-slave-06 ?
> >
> > Olaf
> >
> >