Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 18 Apr 2007 12:57:28 +0400
From:      Yar Tikhiy <yar@comp.chem.msu.su>
To:        Pawel Jakub Dawidek <pjd@freebsd.org>
Cc:        Max Laier <max@love2party.net>, freebsd-arch@freebsd.org
Subject:   Re: Host ID.
Message-ID:  <20070418085727.GB40826@comp.chem.msu.su>
In-Reply-To: <20070409163456.GJ76673@garage.freebsd.pl>
References:  <20070407120656.GD63916@garage.freebsd.pl> <20070409145620.GF76673@garage.freebsd.pl> <200704091601.27454.max@love2party.net> <20070409163456.GJ76673@garage.freebsd.pl>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Apr 09, 2007 at 06:34:56PM +0200, Pawel Jakub Dawidek wrote:
> On Mon, Apr 09, 2007 at 04:01:22PM +0200, Max Laier wrote:
> > On Monday 09 April 2007 16:56, Pawel Jakub Dawidek wrote:
> > > On Sat, Apr 07, 2007 at 02:06:56PM +0200, Pawel Jakub Dawidek wrote:
> > > > Hi.
> > > >
> > > > After initial discussion on IRC, I'd like to propose an addition...
> > > > I want to use it with ZFS, but I thought it may be useful in general,
> > > > so here it goes:
> > > >
> > > > I'd like to assign a unique ID to the system on first boot.
> > > >
> > > > When system starts, /etc/rc.d/hostid script checks if /hostid file
> > > > exists, if it doesn't, it creates it via 'uuidgen > /hostid'.
> > > >
> > > > It will also set kern.hostuuid sysctl to this value and first four
> > > > bytes of MD5(kern.hostuuid) will be stored in kern.hostid. It will
> > > > allow to use gethostid(3).
> > > >
> > > > If root file system is read-only, different uuid will be genrated on
> > > > each boot. Not sure if anything better can be done here.
> > > >
> > > > As I said, I think it may be genrally useful. Imagine using it with
> > > > magic/variant symlinks, for example.
> > >
> > > Here is the patch:
> > >
> > > 	http://people.freebsd.org/~pjd/patches/hostid.patch
> > >
> > > Any objections?
> > 
> > I suppose:
> > 
> > > Index: usr.bin/uuidgen/Makefile
> > > ...
> > > +BINDIR=	/bin
> > 
> > will be fixed by (repo)copy for the real commit?  Otherwise, no.
> 
> kan suggested not to do it: "respect CVS handicaps, please...." :)

It can be repocopied some day in future, after all the stuff settles.
Perhaps we should grep src/usr.bin for moved pieces and submit a
collective usr.bin->bin repocopy request to the repomeisters.  I
have the repocopy of pkill in my personal list, which has been happy
in /bin for quite a while now but still has its sources in usr.bin.

-- 
Yar



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070418085727.GB40826>