Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!munnari.oz.au!metro!news
From: szabo_p@maths.su.oz.au (Paul Szabo)
Newsgroups: comp.sys.apollo
Subject: Re: Booting question on an DN10K
Message-ID: <1990Aug13.010057.24081@metro.ucc.su.OZ.AU>
Date: 13 Aug 90 01:00:57 GMT
References: <573@synopsys.COM>
Sender: szabo_p@maths.su.oz.au (Paul Szabo)
Reply-To: szabo_p@maths.su.oz.au (Paul Szabo)
Organization: Dept of Applied Mathematics, University of Sydney
Lines: 25

In article <573@synopsys.COM>, bala@synopsys.com writes:
>We have a DN10k running Sr10.2p off its local disk. We want to add another
>local disk to it and load SR10.0.2p on it. Then we would like to be able boot 
>off the 2nd disk (under Sr10.0.2) OR the 1st disk (under Sr10.2p) on demand. 
>
>Thanks to this group, we were able to do similar stuff on a DN4500, by setting
>the boot volume using the 'DI w0:1' command. But, is the same thing possible
>on a DN10K?

I am not sure if 'DI w0:1' would work on the DN10000, probably it would.
But...

The 'ES EE_CONFIG' programs are different between 10.0.p (we never had
10.0.2.p) and 10.2.p. They are different enough, so that you would need
to run this between boots of the 'other' OS, even though the
configuration of your node does not change. This is documented in the
10.1.p release notes (Appendix B).

Also, if you have a 'Floppy/Cartridge tape (SCSI) controller' board
(part no. 12171), this needs to be re-jumpered between 10.0.p and
10.2.p. Yes, this sounds rather odd, but it is true: if you have the
board jumpered for the wrong OS release, you will not be able to boot.


Paul Szabo      szabo_p@maths.su.oz.au