Date: 31 Oct 2002 23:01:22 +0000 From: Wayne Pascoe <freebsd@penguinpowered.org.uk> To: Steve Warwick <ukla@attbi.com> Cc: <freebsd-questions@FreeBSD.ORG> Subject: Re: Separating the OS from the data [Addendum] Message-ID: <86bs5aciwt.fsf@marvin.penguinpowered.org.uk> In-Reply-To: <B9E6F0A0.3B38%ukla@attbi.com> References: <B9E6F0A0.3B38%ukla@attbi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Steve Warwick <ukla@attbi.com> writes:
> [Addendum]
>
> Cvsup / makeworld: I apologize for missing that piece of information
>
> Yes, I could use the usual update procedure, however, this is a production
> machine. So my thought is: build a new OS on a staging machine, add required
> symlinks, pull the drive (sled) and slot it into the production machine. In
> THEORY it should be possible to do an upgrade in the time it takes to do a
> reboot. For server farms this would be a big benefit...
Still not an answer, but something that may help...
We have a test box that mirrors the hardware and software conf of our
live farm. We do the make world on the staging box, then NFS export
/usr/src and /usr/obj to each of our production machines.
We then just do a make installworld on each of the production
machines, copy the new kernel into place and do a reboot.
We have 14 machines and we've been doing this since 4.2 with no
hassles.
--
- Wayne Pascoe
Everything to excess. To enjoy the flavour of
life, take big bites. Moderation is for
monks. - Robert Heinlein
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?86bs5aciwt.fsf>