[3533] in SIPB_Linux_Development
Re: Problems with updating recently loaded machines
daemon@ATHENA.MIT.EDU (Angie Kelic)
Mon Aug 13 23:25:34 2001
Message-Id: <200108140325.XAA21764@airfoil.mit.edu>
To: cjt@lcs.MIT.EDU
Cc: linux-dev@MIT.EDU
Reply-To: linux-dev@MIT.EDU
In-reply-to: Your message of "13 Aug 2001 23:13:41 EDT."
<sjm1ymf72qi.fsf@rcn.ihtfp.org>
Date: Mon, 13 Aug 2001 23:25:31 -0400
From: Angie Kelic <sly@MIT.EDU>
Hi there,
I am running SMP machines under 9.0. None of the updates
correctly deal with updating the SMP kernel. You need to
do that part by hand (but it does work and will interoperate
just fine with athena). I would recommend following the kernel
update instructions at:
http://www.redhat.com/support/docs/howto/kernel-upgrade/kernel-upgrade.html
Essentially you need to install the kernel-smp package
along with the standard kernel package of the same version,
run mkinitrd for the smp kernel if applicable and reconfigure
and rerun lilo. AFS and all other things should work correctly.
This is the case for both the 8.4.25 and the 9.0 releases (note
that you should make sure you have the correct version of kernel-smp
-- for 8.4.25 I believe it's 2.2.17 and for 9.0 it's 2.4.3-12 --
you'll find these rpms in
/afs/athena/system/rhlinux/redhat-6.2-updates/RPMS
and
/afs/athena/system/rhlinux/redhat-7.1-updates/RPMS respectively).
I"m not sure why you're running into the yp-bind conflict,
The easiest way to resolve this may be to do an
rpm -e ypbind yptools
before you run the update -- of course if you are actually
using an NIS domain this may cause problems for you.
Good luck.
--Angie
>Yes, SMP is known not-to-work in the current version.
>
>-derek
>
>Chris Terman <cjt@lcs.mit.edu> writes:
>
>> [If there's a better place to send this info please let me
>> know -- Thanks.]
>>
>> After loading several different machines in the 6.004 lab
>> last week using the SIPB-IS linux-boot-floppy, running
>> update_ws (either automatically or manually) to version
>> 8.4.25 causes the kernel-smp package to be erased. This
>> seems to break detection of the eth0 interface on reboot and
>> subsequently the ifup script in sysconfig/network-scripts fails
>> since eth0 is not found by ifconfig. Both machines (an HP XU6/200
>> and a Dell Optiplex GX1) actually had dual-processor motherboards,
>> and the Dell had a second processor installed. However, the
>> vanilla 6.001 workstations seem have to suffered the same failure.
>>
>> Attempts to run "update_ws 9.0" produces the following
>> conflict report:
>>
>> Update would break dependencies:
>> yp-tools is needed by ypbind-1.7-0.6.x
>> *** The update has failed ***
>>
>> Neither problem is crucial, but it would be nice to be able to
>> keep the 6.004 lab machines on autoupdate and to eventually
>> migrate to the 2.4 kernel, so any hints as to workarounds would
>> be helpful. Thanks.
>>
>> - Chris
>>
>>
>
>--
> Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
> Member, MIT Student Information Processing Board (SIPB)
> URL: http://web.mit.edu/warlord/ PP-ASEL-IA N1NWH
> warlord@MIT.EDU PGP key available
