*BSD News Article 3448


Return to BSD News archive

Path: sserve!manuel!munnari.oz.au!mips!mips!darwin.sura.net!gatech!bloom-beacon!eru.mt.luth.se!lunic!sunic!aun.uninett.no!barsoom!barsoom!tih
From: tih@barsoom.nhh.no (Tom Ivar Helbekkmo)
Newsgroups: comp.unix.bsd
Subject: Re: 2nd drive file system destroyed on re-boot
Message-ID: <tih.713450893@barsoom>
Date: 10 Aug 92 12:48:13 GMT
References: <Bsp5D6.B9p@obiwan.uucp> <165f2uINN6an@disaster.Germany.EU.net>
Sender: news@barsoom.nhh.no (USENET News System)
Organization: Norwegian School of Economics
Lines: 45

In article <Bsp5D6.B9p@obiwan.uucp>, bob@obiwan.uucp (Bob Willcox) writes:

> I have discovered that specification of the sf key (bad sector forwarding)
> in the disktab entry for my second disk results in distruction of the file
> system on that disk.

There's more to it than this, I think...  I've modified the wd driver
quite a bit by now, and have it working very well indeed with two
controllers and 4 physical disk drives.  Since the two controllers can
run in parallell, distributing the load carefully over the disks gives
quite good results.  Heavy disk thrashing causes no troubles at all.
However, for this scheme to work, I have to run the drives on the 2nd
controller without the bad144 style bad sector forwarding.

If, for instance, I run "disklabel -e -r wd3" and add the "badsect"
flag to the disklabel, I can thereafter do a "bad144 wd3 0", and
things will work just fine -- until the next reboot.  As soon as the
disk driver tries to read the bad sector table over at the end of the
disk, it starts recalibrating (I think) like nobody's business.  In
fact, it gets so enthusiastic about shoving the heads back and forth
that it managed to physically destroy an old Seagate drive I had
hooked up to it!  :-)

This is with a warm reboot, however.  If I turn the machine off, then
on, things look a bit better: I can read the disk label, which shows
the badsect flag, but if I run "bad144 wd3" to display the bad sector
table, it turns out that it's all corrupt.  I can use disklabel to
remove the badsect flag, and I'll be OK again.

bs@Germany.EU.net (Bernard Steiner) writes:

>I noticed that with my ESDI drives I had to give disklabel one cylinder less
>than what the physical drive parameters are...

I thought of this too, but I've tried reducing the number of available
cylinders to about a fourth of the real number with no improvement...

I'm still debugging this stuff, prior to releasing the driver in its
final form, so if anyone has any kind of input at all regarding
problems with the sector forwarding code, please let me know!

-tih
--
Tom Ivar Helbekkmo, NHH, Bergen, Norway.  Telephone: +47-5-959205
Postmaster for domain nhh.no.   Internet mail: tih@barsoom.nhh.no