RE: Serious problems on DS20

["Champigny, Michael" <Michael Champigny compaq com>]

By the way, this was also the root of a problem I had installing
Linux on a Jensen box. In this case I was using minlabel, which
has the same bug. If I set the last cylinder to it's true maximum,
I ended up corrupting my first cylinder.  I had to back it off by
1 in order to avoid the wraparound. Then I could install normally.

/Michael

-----Original Message-----
From: Jemiolo, John 
Sent: Thursday, October 21, 1999 3:32 PM
To: 'axp-list@redhat.com'
Subject: RE: Serious problems on DS20


Brad
Did you use FDISK to partition your disk?, if you did, did you enter
maxcyl-1 as the last cylinder to use. FDISK has a nasty habit of wrapping
the last cylinder, which does lead to file system corruption.

JJ

-----Original Message-----
From: Brad Lucier [mailto:lucier@math.purdue.edu]
Sent: Thursday, October 21, 1999 3:24 PM
To: axp-list@redhat.com
Subject: Re: Serious problems on DS20


I've been trying to get a dual-processor EV6 (Tsunami board, DS20
clone) to run RH Linux properly since the beginning of August, with
no luck.  At this point, I'm ready to try nearly anything--- a different
distribution, new controllers, new peripherals, anything.

Somebody *please* tell me a success story with a configuration like mine.
The only constraint is that it has to run SMP.

Brad Lucier    lucier@math.purdue.edu

Latest installation (10/15/99)

Clean RH 6.0 install from CD.
Installed all 6.0 update RPMs except for the kde updates.
Fixed, built, and installed binutils-2.9.5.0.16.
Built and installed new kernel 2.2.13pre17 with the following non-default
options from make menuconfig:
CONFIG_SMP=y
CONFIG_SCSI_AIC7XXX=y
CONFIG_AIC7XXX_CMDS_PER_DEVICE=8 (default)
CONFIG_AIC7XXX_RESET_DELAY=5 (default)
CONFIG_SCSI_NCR53C8XX=y
CONFIG_SCSI_NCR53C8XX_DEFAULT_TAGS=8 (default)
CONFIG_SCSI_NCR53C8XX_MAX_TAGS=32 (default)
CONFIG_SCSI_NCR53C8XX_SYNC=20 (default)
CONFIG_DE4X5=y

Latest failure mode: (10/19/99)

Starting trashing the root partition when gzipping a large file as
a normal user and started untarring linux-2.2.12.tar.gz as root.
(Was going to build a proper 2.2.13 kernel from 2.2.12+patch.)
Trashed the entire disk so badly another clean install is needed.

Installed Hardware: (from "show config" on the SRM console)

SRM Console: V5.4-19
PALcode: OpenVMS PALcode v1.54-43, Tru64 UNIX PALcode v1.49-45
PCI Hose 00
    Bus 00 Slot 06/0: Adaptec AIC-7895 pkb0.7.0.6.0 SCSI Bus ID 7
    Bus 00 Slot 06/1: Adaptec AIC-7895 pkc0.7.0.106.0 SCSI Bus ID 7
    Bus 00 Slot 09/1: NCR53c810 pkd0.7.0.9.0 SCSI Bus ID 7
                                dkd600.6.0.9.0 Toshiba CD-ROM XM-64
PCI Hose 01
    Bus 00 Slot 09:   NCR53c895 pk0.7.0.7.1 SCSI Bus ID 7
                               dka0.0.0.7.1 IBM DDRS-39130D

How I did the install:

Attached CD and disk to Adaptec controllers, did install; upgraded
rpms; built and installed binutils-2.9.5.0.16; built and installed
2.2.13pre17 with egcs-1.1.2 and the new binutils; moved CD and disk
to NCR controller (on-board Adaptec contollers suck big-time).

-- 
To unsubscribe: send e-mail to axp-list-request@redhat.com with
'unsubscribe' as the subject.  Do not send it to axp-list@redhat.com

-- 
To unsubscribe: send e-mail to axp-list-request@redhat.com with
'unsubscribe' as the subject.  Do not send it to axp-list@redhat.com