*BSD News Article 84901


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!feed1.news.erols.com!howland.erols.net!news3.cac.psu.edu!news.math.psu.edu!scramble.lm.com!not-for-mail
From: Michael Holling <myke@lm.com>
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Hard error writing to disk?
Date: 13 Dec 1996 06:44:14 GMT
Organization: Telerama Public Access Internet, Pittsburgh, PA
Lines: 26
Message-ID: <58qtvu$4ok$1@scramble.lm.com>
NNTP-Posting-Host: berzerk.lm.com
X-Newsreader: TIN [UNIX 1.3 unoff BETA release 960917]
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.freebsd.misc:32518

I'm running 2.1R on a 486 with two drives.  wd2 is on the secondary controller
and holds /usr.  Today I encountered a situation where trying to write a file
to a certain directory would cause the same error each time.  I was able to
write the same file to another spot on the disk without incident.  After 
writing about half the file I started getting errors like this:

wd2s2e: hard error writing fsbn xxx of yyy - zzz (wd2s2 bn xxx; cn 376 tn 47
sn 31)wd2: status 61 <rdy,wrtflt,err> error 4<abort>

(xxx,yyy,zzz are actual large numbers, I didn't write them down)

At this point the system would crash hard, the vm system started giving errors
and everything went rapidly downhill from there.  Sometimes I was able to
halt the system (it couldn't sync the disks however), othertimes I had to do
hard reboots.  A fsck on /dev/rwd2s2e comes up clean every time.  It's been a
while since I installed this system but I'm almost positive I did a bad block
scan on wd2 (is there any way to check?).  Also, half the time the system
would reboot only to have the BIOS report that the secondary HD controller
was not working, powercycling fixes this.

Any hardware gurus know offhand where the problem might lie?  Or any other 
diagnostic software to pinpoint the error?  It is 100% reproducible, but I'm
not really sure where to go from here...

-- 
- Mike