*BSD News Article 53568


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!zombie.ncsc.mil!news.mathworks.com!tank.news.pipex.net!pipex!dispatch.news.demon.net!demon!kiss.demon.co.uk
From: phil@zipmail.co.uk (Phil Taylor)
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Problem With Fujitsu SCSI Drive
Date: Wed, 25 Oct 1995 21:55:11 GMT
Organization: Lan Systems
Lines: 62
Message-ID: <814658177.16718@kiss.demon.co.uk>
Reply-To: phil@zipmail.co.uk
NNTP-Posting-Host: kiss.demon.co.uk
X-NNTP-Posting-Host: kiss.demon.co.uk
X-Newsreader: Forte Free Agent 1.0.82

I am trying to set up FBSD 2.0.5R on a system with an Adaptec AHA1542B
and a Fujitsu M2694ESA 1GB SCSI-2 Drive.

It isn't working ......

Everytime I run the installation it works fine and goes through
without any problem !!!!

As the drive had been recently low-level formatted (with scsifmt) it
was completely blank and fdisk found its geometry as 1800 odd cyls
with a stupid number of hds/sects and it moaned about invalid
geometry. I created a partition with dos fdisk and looked at the
partition table and it was 

1023 cyls 63 hds and 32 secs.

When I ran fbsd install/fdisk again this time the geometry was
detected as:

1033 cyls 64 hds and 32 secs. !!!!
  ^^       ^

The first time I left it at this and finished the install, after a few
reboots/power-cycles when fsck was running the drive seemed to be
hitting the end-stop and making some 'very funny noises'.

I got a lot of very nasty looking HARDWARE FAILURE errors and quite a
few vm - probable hardware fault and the occasional help me message !!

I can post the actual messages if required but I think you get the
general idea, it was not happy 8-(

I didn't think that the scsi drive would allow fbsd to tell it to go
to an area of disk that doesn't exist but I thought I would try what
DOS had suggested 1023/63/32.

This seemed to work fine then suddenly off it went again.

You are probably all now saying 'F****D DRIVE' but I have tested it
under dos with norton calibrate and other tools and it was OK. I am
going to run novell's compsurf tonight and will post the results. Now
I know that unixes tend to be a lot more stressful on hardware (and
sometimes sanity) but for it to not fail under dos seems strange if it
were such a hardware fault.

It is not the controller as I have tried a 1542CF as well.

It is properly terminated and I have tried two other SCSI cables
(different types)

I have tried sync and async and a variety of DMA speeds ......

If someone has seen this before and proven it to be either hardware or
configuration related then I would love to hear from you, time is at a
premium and I need to know if I have to scrap this drive and buy a new
one pretty soon !!!!!!!

Cheers 

Phil