Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!ibmsupt.uucp!slo!jsalter
From: jsalter@slo.uucp (James Salter)
Newsgroups: comp.unix.aix
Subject: Re: installing several AIX PS/2's
Keywords: cloning a system
Message-ID: <1990Jun27.173958.15364@ibmpa>
Date: 27 Jun 90 17:39:58 GMT
References: <1990Jun25.154509.6726@cunixf.cc.columbia.edu> <4432@milton.u.washington.edu> <18418@rpp386.cactus.org> <9006270009.AA14366@mindcrf.mindcraft.com>
Sender: news@ibmpa (news id)
Reply-To: jsalter@slo.UUCP (James Salter)
Organization: IBM AWD, Palo Alto
Lines: 35

In article <9006270009.AA14366@mindcrf.mindcraft.com> karish@mindcrf.UUCP writes:
>In article <18418@rpp386.cactus.org> jfh@rpp386.cactus.org
>(John F. Haugh II) writes:
>>In article <4432@milton.u.washington.edu> eliot@milton.u.washington.edu
>(Eliot Lim) writes:
>There's actually an apparently-supported procedure to do this,
>documented in the release notes I got with 9013S.  The sequence is
>  cd /
>  mkszfile
>  mksysb <tape device>
>  Boot the second system from diskettes, and install from the tape
>  just made.

The proper procedure for this is to type (as root):

	smit startup

which will give you a smit menu with a backup option.  This options allows
backing up of the relevant information for cloning the current system
configuration.

>Failure to heed these warnings may result in the generation of a very
>confused ODM database.  If all your systems are the same, though, and
>each has a big disk, it's easy and safe.

Do NOT copy ODM databases across machines.  Because the ODM uses physical 
ID values of disks for some things, there may be problems if you attempt to
use one machine's ODM databases (/etc/objrepos/Cu*) on a different machine.

>	Chuck Karish		karish@mindcraft.com

jim/jsalter  IBM AWD, Palo Alto  T465/(415)855-4427    VNET: JSALTER at AUSVMQ
Internet: ibmsupt!jsalter@uunet.uu.net         UUCP: ..!uunet!ibmsupt!jsalter 
  "Life can be modelled as a hidden Markov process with infinite states and
   no a-priori knowledge of the probability density functions." - dakramer