Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 7 Aug 2002 20:32:55 +0200
From:      Erik Trulsson <ertr1013@student.uu.se>
To:        Christopher Schulte <schulte+freebsd@nospam.schulte.org>
Cc:        Josh Paetzel <friar_josh@webwarrior.net>, BSD Freak <bsd-freak@mbox.com.au>, "Roger 'Rocky' Vetterberg" <listsub@401.cx>, FreeBSD Questions <freebsd-questions@FreeBSD.ORG>
Subject:   Re: There must be a better way to maintain older systems
Message-ID:  <20020807183254.GA68082@falcon.midgard.homeip.net>
In-Reply-To: <5.1.1.6.2.20020807112827.05fde6f8@localhost>
References:  <e5fe64e5c22e.e5c22ee5fe64@mbox.com.au> <e5fe64e5c22e.e5c22ee5fe64@mbox.com.au> <5.1.1.6.2.20020807112827.05fde6f8@localhost>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Aug 07, 2002 at 11:46:17AM -0500, Christopher Schulte wrote:
> At 11:16 AM 8/7/2002 +0000, Josh Paetzel wrote:
> >It's already been suggested that you build a machine specifically
> >for doing buildworlds on and then installing all the other machines
> >off the first one.  Total downtime for production server == one
> >reboot to load the new kernel.
> 
> One curiosity I've considered with this method.
> 
> If I buildworld on a generic box with a generic make.conf,
> will the installworld on target machine with mounted /usr/src and
> /usr/obj follow its local version of make.conf with install?  Such as,
> if I don't want to install bind on one specific target machine, will
> 
> -- /etc/make.conf --
> NO_BIND=        true
> -- /etc/make.conf --
> 
> on target prevent bind from being installed with an installworld
> even though it was built with buildworld on the build server?

That should work just fine.  Just be careful so you don't try to
install things you didn't build. (I.e. having NO_BIND=true on the
*build*machine while having NO_BIND commented out on the install
machine is a bad idea.)


> 
> If so, I could generate a daily generic 'build' as well as custom
> kernels for all of my servers automatically after my local cvsup
> server pulled an update from cvsupxx.freebsd.org.  Then as an update
> was required on any specific server, it would be a matter of mounting
> the build directories, and taking it from there.  World and kernel would
> be all fresh and warm.  Perhaps cp -Rp over from the network first
> so I can properly boot into single user mode and not rely on the network
> to finish the install.  Hell, that too could be automated so /usr/src
> and /usr/obj are populated on the local disk of every machine.
> 
> Some sanity checks would need to be done before issuing the
> actual update commands, but I could see this method working.
> 
> Comments?

This method should be perfectly workable.  I can't think of anything in
particular that would prevent it from working.

It would probably be a good idea to test the new build on some
unimportant machine before installing it on all machines but I assume
that is included in the "sanity checks" mentioned above.

-- 
<Insert your favourite quote here.>
Erik Trulsson
ertr1013@student.uu.se

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?20020807183254.GA68082>