What to do about Infrastructure?

[Jarod Wilson]

On Sat, Aug 21, 2010 at 8:02 AM, Thorsten Leemhuis <fedora at leemhuis.info> wrote:
> Hi!
>
> There yet again was a serious misconfiguration on our one and only
> x86-{32,64} builder that was caused indirectly by the branching of
> Fedora 14 a week ago. That's the reason why there are some packages with
> the tag "fc15" in our rawhide repos, which are still targeting F14, as
> we haven't branched yet. It's should be fixed now; thx Nicolas for
> letting me know (and partly blame me, as I should have seen that earlier
> and checked things when the adjustments for the builder were discussed
> something like two weeks ago iirc)
>
> That IMHO is just another(¹) sign that our infrastructure team doesn't
> work well (if at all). I don't blame anyone, but we IMHO really need to
> do something about it. A full reboot / fresh start with new people seems
> way overdue to many.
>
> Are there any options on it? Or maybe anybody even willing to lead that
> effort?

I've tried to get more involved a few times, and its never really
managed to go anywhere. I'd like to do more than just provide a build
host, and occasionally hop on and fix (or break) things on it, but the
reality is that I probably don't have the time to do much more than
that.

That said, I'm just about ready to replace the vm I currently provide
as a builder with real hardware. I've got a glut of atom 330 boxes
right now, so I was about to roll out a dual-core atom 330 box with
4GB of RAM and 1.5TB of storage (2x 250G internal + 1TB e-sata) as a
replacement for the vm. It should be capable of providing additional
services, if desired. (I've got other slightly more powerful hardware
that could be rolled out as well, but went with this system for power
consumption considerations).

-- 
Jarod Wilson
jarod at wilsonet.com