----- Original Message -----
From: "Mike Burns" <mburns(a)redhat.com>
To: "Itamar Heim" <iheim(a)redhat.com>
Cc: "Eyal Edri" <eedri(a)redhat.com>, infra(a)ovirt.org
Sent: Thursday, July 12, 2012 6:51:00 PM
Subject: Re: Infrastructure for Jenkins
On Thu, 2012-07-12 at 17:57 +0300, Itamar Heim wrote:
> On 07/12/2012 05:55 PM, Eyal Edri wrote:
> >
> >
> > ----- Original Message -----
> >> From: "Karsten 'quaid' Wade" <kwade(a)redhat.com>
> >> To: infra(a)ovirt.org
> >> Sent: Thursday, July 12, 2012 5:31:41 PM
> >> Subject: Re: Infrastructure for Jenkins
> >>
> >> -----BEGIN PGP SIGNED MESSAGE-----
> >> Hash: SHA1
> >>
> >> On 07/12/2012 05:57 AM, Jabs, Joachim wrote:
> >>> Hi Guys,
> >>>
> >>>
> >>>
> >>> as previously offered:
> >>>
> >>>
> >>>
> >>> Im willing to provide you guys infrastructure for Jenkins. Im
> >>> currently trying out getting oVirt running on a cluster (we
> >>> have
> >>> 192 Cores and 3TB RAM aviable) for providing help with problems
> >>> around ovirt that might arise.
> >>>
> >>> The Cluster is not commercially used (Not right now) and for me
> >>> its
> >>> also a good opportunity to learn about ovirt and also about
> >>> hosting
> >>> VM related infrastructure. This will also be a test for
> >>> stability
> >>> and maintainability around the hardware used.
> >>>
> >>>
> >>>
> >>> If you guys sum up what you need, I think I might be able to do
> >>> something for you there. If you have any questions regarding
> >>> the
> >>> offer, just ask.
> >>
> >> Hey, thanks for following. This is a great offer, +1 from me.
> >>
> >> Ewoud, Eyal (& others?) - what does Joachim need to do? Is this
> >> a
> >> good
> >> place to start? --
> >
> > This is an excellent offer and opportunity.
> > Since we're talking about VMs, we can start with basic hardware
> > spec
> > and increase it after we'll see the bottle necks in real time.
> >
> > I would start with 3 VMs each running 16GB RAM, 200GB DISK, 32 GB
> > SWAP, each with 4/8 cores each.
> >
> > Itamar, mburns -> any recommendation on how many VMs per hardware
> > node should we use?
> >
>
> considering the workload, i'd start with not overcomitting
> resources,
> and later analyzing behavior
Agree with Itamar. I'd also consider having a vm or 2 that are
running
various target distros where we can build packages for different
distros.
Also, 200GB disk seems a bit excessive to me. Since this is running
on
oVirt, maybe a thin provisioned disk would be a good choice. The
slaves
generally don't need as much disk.
agreed, especially if it's thin provisioned.
amount of disk space will be determine if we'll want to run multiple jobs on each
slave so it will require cloning the workspaces,
but of course slaves usually require much less space than the master node.
Mike
>
> >
> >>
> >>
http://ovirt.org/wiki/Jenkins
> >>
> >> Cheers - Karsten
> >> - --
> >> Karsten 'quaid' Wade, Sr. Analyst - Community Growth
> >>
http://TheOpenSourceWay.org .^\
http://community.redhat.com
> >> @quaid (identi.ca/twitter/IRC) \v' gpg: AD0E0C41
> >>
> >>
> >> -----BEGIN PGP SIGNATURE-----
> >> Version: GnuPG v1.4.12 (GNU/Linux)
> >> Comment: Using GnuPG with Mozilla -
http://enigmail.mozdev.org/
> >>
> >> iD8DBQFP/t/N2ZIOBq0ODEERAkM+AKCqsLii5DeVVDcueUR/0075fmSWhACfbmGP
> >> PMWT116DTADxn/oNn/pXav4=
> >> =EarP
> >> -----END PGP SIGNATURE-----
> >>
_______________________________________________
> >> Infra mailing list
> >> Infra(a)ovirt.org
> >>
http://lists.ovirt.org/mailman/listinfo/infra
> >>
>
>
> _______________________________________________
> Infra mailing list
> Infra(a)ovirt.org
>
http://lists.ovirt.org/mailman/listinfo/infra