El Mon, 27 Feb 2012 18:03:01 -0500
M Abed <maxamaxim(a)me.com> escribió:
Does that mean that we will remove the following machines from the
heavybuilder channel or leave those as is?
cdot-panda-10-1-v7hl
cdot-panda-10-4-v7hl
cdot-panda-10-5
cdot-panda-11-4-v7hl
cdot-panda-12-3
cdot-panda-5-3
cdot-panda-5-4
It would mean they are no longer heavybuilders. we could then drop the
space thats allocated to them in the scratch pool to be smaller.
Dennis
Masihul Max Abed
Seneca CDOT
On Mon, Feb 27, 2012 at 5:36 PM, Brendan Conoboy <blc(a)redhat.com>
wrote:
> On 02/27/2012 02:25 PM, Dennis Gilmore wrote:
> [snip]
>
> I wanted to get others thoughs before i went and made the change
>>
>
> Just to be clear, these two systems are the same machines:
>
> hsv-trimslice-9-v5tel N N 0.0/2.0 arm -
> hsv-trimslice-9-v7hl Y N 2.0/2.0 armhfp 2012-02-27
> 22:04:45
>
> We just provision them as v5 or v7 according to what it looks like
> is needed. That's the idea anyway- we aren't doing enough builds
> at once that we've had to balance anything.
>
> FYI, the HSV pandas also have dedicated sata disks on them and are
> roughly the performance equivalent of the trimslices because of
> it. The actual tally of heavybuilders (defined as having dedicated
> usb-sata storage) is:
>
> hsv-panda-1-v5tel
> hsv-panda-2-v5tel
> hsv-panda-3-v5tel
> hsv-panda-5-v5tel
> hsv-panda-6-v5tel
> hsv-panda-8-v5tel
> cdot-trimslice-13-1
> cdot-trimslice-14-1-v7hl
> hsv-trimslice-10-{v5tel,v7hl}
> hsv-trimslice-9-{v5tel,v7hl}
> hsv-trimslice-8-{v5tel,v7hl}
> hsv-trimslice-7-{v5tel,v7hl}
> hsv-trimslice-6-{v5tel,v7hl}
>
> We can make all the HSV trimslices into v7 builders and that will
> give us a 6/6 split, which seems reasonable.
>
> --
> Brendan Conoboy / Red Hat, Inc. / blc(a)redhat.com
>
> ______________________________**_________________
> arm mailing list
> arm(a)lists.fedoraproject.org
>
https://admin.fedoraproject.**org/mailman/listinfo/arm<https://admin.f...
>