You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by Gavin McDonald <ga...@16degrees.com.au> on 2016/11/09 03:03:33 UTC
Jenkins Lucene Node
Hi Guys,
You have one executor on one node (lucene label) which is on average having 5 or more (up to 12 usually) jobs waiting in the queue.
As we are striving to improve our services overall and Jenkins is getting quite a bit of attention lately, we feel that Lucene is in need
of more resources.
With that in mind, we’d like to offer another node for your exclusive use - a duplicate of the current node, so you can spread your jobs
across 2 executors (just by applying the lucene label between the 2 nodes.)
WDYT? If yes, let me know and I’ll get it started. If no, also do please me know.
Thanks
Gav… (ASF Infra)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org
RE: Jenkins Lucene Node
Posted by Uwe Schindler <uw...@thetaphi.de>.
Hi Gav,
it would be nice to get more exceutors, yes. The main reason why there are so many runs in queue is just for a different reason:
As Lucene/Solr runs randomized tests, we have to run them all the time. So every job should run as often as possible. Our builds are not triggered by something like git commits. So we have a cronjob on every job that just triggers a new build every day or every hour (@daily, @hourly). This makes the tests run all the time. So if all is fine you would see a build of every job in the queue. Jenkins never queues the same job multiple times, so the trick is just to keep the build queue full with one build of each job.
So when we have a second machine, the number of waiting builds will not really change, but the throughput would be higher.
Uwe
-----
Uwe Schindler
H.-H.-Meier-Allee 63, D-28213 Bremen
http://www.thetaphi.de
eMail: uwe@thetaphi.de
> -----Original Message-----
> From: Steve Rowe [mailto:sarowe@gmail.com]
> Sent: Wednesday, November 9, 2016 2:49 PM
> To: dev@lucene.apache.org; private@infra.apache.org
> Subject: Re: Jenkins Lucene Node
>
> Hi Gav,
>
> Yes please!
>
> Probably wishful thinking, but: we have several so-called Monster tests that use
> a lot of resources - I run them on my personal Jenkins with a 30GB heap and a
> disk that can accomodate ~1TB of temporary disk usage. Any chance the new
> VM would be capable of running these?
>
> --
> Steve
> www.lucidworks.com
>
> > On Nov 8, 2016, at 10:03 PM, Gavin McDonald <ga...@16degrees.com.au>
> wrote:
> >
> > Hi Guys,
> >
> > You have one executor on one node (lucene label) which is on average having
> 5 or more (up to 12 usually) jobs waiting in the queue.
> >
> > As we are striving to improve our services overall and Jenkins is getting quite
> a bit of attention lately, we feel that Lucene is in need
> > of more resources.
> >
> > With that in mind, we’d like to offer another node for your exclusive use - a
> duplicate of the current node, so you can spread your jobs
> > across 2 executors (just by applying the lucene label between the 2 nodes.)
> >
> > WDYT? If yes, let me know and I’ll get it started. If no, also do please me
> know.
> >
> > Thanks
> >
> > Gav… (ASF Infra)
> >
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
> > For additional commands, e-mail: dev-help@lucene.apache.org
> >
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
> For additional commands, e-mail: dev-help@lucene.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org
Re: Jenkins Lucene Node
Posted by Steve Rowe <sa...@gmail.com>.
Hi Gav,
Yes please!
Probably wishful thinking, but: we have several so-called Monster tests that use a lot of resources - I run them on my personal Jenkins with a 30GB heap and a disk that can accomodate ~1TB of temporary disk usage. Any chance the new VM would be capable of running these?
--
Steve
www.lucidworks.com
> On Nov 8, 2016, at 10:03 PM, Gavin McDonald <ga...@16degrees.com.au> wrote:
>
> Hi Guys,
>
> You have one executor on one node (lucene label) which is on average having 5 or more (up to 12 usually) jobs waiting in the queue.
>
> As we are striving to improve our services overall and Jenkins is getting quite a bit of attention lately, we feel that Lucene is in need
> of more resources.
>
> With that in mind, we’d like to offer another node for your exclusive use - a duplicate of the current node, so you can spread your jobs
> across 2 executors (just by applying the lucene label between the 2 nodes.)
>
> WDYT? If yes, let me know and I’ll get it started. If no, also do please me know.
>
> Thanks
>
> Gav… (ASF Infra)
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
> For additional commands, e-mail: dev-help@lucene.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org