*BSD News Article 9645


Return to BSD News archive

Received: by minnie.vk1xwt.ampr.org with NNTP
	id AA6176 ; Tue, 05 Jan 93 16:18:52 EST
Newsgroups: comp.unix.bsd
Path: sserve!manuel.anu.edu.au!munnari.oz.au!uunet!cs.utexas.edu!uwm.edu!spool.mu.edu!news.cs.indiana.edu!mikes@cs.indiana.edu
From: "Michael Squires" <mikes@cs.indiana.edu>
Subject: Re: [386BSD] corrupt files
Message-ID: <1993Jan8.004152.16521@news.cs.indiana.edu>
Summary: is it an old ESDI disk?
Keywords: ESDI bad144 disktab
Organization: Indiana University, Bloomington
References: <56750001@acf3.NYU.EDU> <1993Jan6.043012.53@kumr.lns.com> <C0HrHK.AtK@agora.rain.com>
Date: Fri, 8 Jan 1993 00:41:42 -0500
Lines: 50

In article <C0HrHK.AtK@agora.rain.com> rgrimes@agora.rain.com (Rodney Grimes) writes:
>pozar@kumr.lns.com (Tim Pozar) writes:
>
>>chapman@acf3.NYU.EDU (Gary W. Chapman, , ,) writes:
>>>After downloading 386BSD from agate.berkeley.edu (and from another place
>>>as well), I get 
>>>
>>>bin01.02 is corrupt
>>>bin01.03 is corrupt

I got that on my 386/20 with an old ESDI drive.  I first tried renaming the
bad files to bad.0? and then getting new copies, but there were too many
bad sectors on the drive and Since the drive is large (660MB) I had to
figure out how to do it using bad144.

So I did the following:

1.  obtain a copy of the kernel that handles WD (WD1007V in my case) errors
    correctly

    I did this by using a second system to install bin01, src01, etc01, and
    the patchkit and then compiling a new kernel.

2.  Figure out a disktab entry for the disk, with bad144 remapping turned on.

    There is info in the unofficial directory from cgd on writing disktab
    entries.  I guessed on the number of cylinders to leave at the end for
    housekeeping.

3.  Make a fixit.fs disk with the new kernel and bads144 from rafal@ref.tfs.com
    which will scan a disk, disklabel the disk, newfs the partitions, and
    rund bads144 to create a list of bad sectors that are fed to bad144 for
    remapping.

    (I did not know at first that once the system is running errors are
    relative to the partition, not absolute sector numbers).

4.  Followed cgd's instructions on creating a 386BSD system using wd0a for
    root, wd0b for swap, wd0h for /usr, and wd0g for /usr/home.

The disk turns out to be quite reliable once the 25 or so bad sectors were
mapped out.

Given what I know now I would head straight for a SCSI controller/disk if
starting from scratch.

-- 

Mike Squires (mikes@iuvax.cs.indiana.edu)     812 855 3974 (w) 812 333 6564 (h)
mikes@iuvax.cs.indiana.edu          546 N Park Ridge Rd., Bloomington, IN 47408