SUMMARY: SMD disk gone bad

From: Mike Jewison (mjewison@nereid.sal.ists.ca)
Date: Fri Jun 25 1993 - 08:42:56 CDT


Earlier this week I had asked:

>We've had one of our SMD disks go bad.
>
>System configuration:
>
> Sun 4/280, SunOS 4.1.3
> Xylogics 7053 SMD disk controller
> Hitachi DK815-10 disk drive
>
>The disk is exhibiting a multitude of errors which I am unable to cure using
>FORMAT. When I get to the ANALYZE menu and run WRITE I am met with the
>following error message:
>
>> Block 522692 (520/1/25), Fatal media error (header not found)
>> Repairing...Warning: unable to save defective data for blkno 522692.
>> Bad block map table overflow.
>> failed.
>
>In fact, I get literally thousands of these error messages for numerous sectors
>on the disk. I checked the System Administration manual and it says that this
>error can be caused (a) by running a surface analysis with a bad cable and
>(b) generally only if the current disk is not configured for slip sectoring.
>
>I have checked and reseated all the cables and the problem will not disappear.
>The fact that the errors occur at the same places on the disk each time an
>analysis is performed leads me to believe that the errors are indeed real.
>
>My questions:
>
>1) Since the disk is no longer on a maintenance contract, we will not be paying
> to get it fixed (or replaced). Am I likely to be able to get it working
> again using format? And do I need to configure the disk for slip sectoring?
>
>2) How *DOES* one configure the disk for slip sectoring? None of the System
> Administration, 7053, or disk drive manuals explain how to do this although
> the 7053 manual explains what slip sectoring is.
>
>3) Is the disk worth the effort of repairing? I'd hate to get it back online
> only to find it's gone flaky again in another couple of weeks.
>
>I should point out we don't *NEED* the disk (otherwise it would be on a
>maintenance contract), but we would (obviously) prefer to have the extra online
>storage the disk provides.
>
>Any insights gratefully accepted. Summary to follow. Many thanks.
>

------------------------------------------------------------------------------

The consensus was don't spend money to repair it but it can't hurt to play
around and see if it will return to life. Several people suggested I format
the drive; I'd tried that previously and although it helped somewhat it didn't
eliminate the problem. Others reported similar problems which went away after
the drive was replaced; as I said in my original post this is not an option
for us. Many respondents said "Forget it, it's gone. Buy a SCSI disk instead."

John Valdes sent me a couple of enclosures, one of which was a C routine which
will tell you the current configuration of your SMD drive. I have made it
available on our anonymous ftp site (nereid.sal.ists.ca) in /pub/smd/smdinfo.c.
If you're interested in getting a copy you can retrieve it from there.

Terry White told me:

>If the disk came from Sun it is configured for slip sectoring.
>Slip sectoring is configured with the dip switches on the drive.
>The dip switches are located on the front panel below the mode
>switch. The switches should be configured as follows,
>
>
>S-27 1off, 2off, 3off, 4on, 5on, 6off, 7on, 8off
>S-41 1off, 2on, 3off, 4off, 5off, 6off, 7off, 8off
>S-28 1off, 2off, 3off, 4off, 5on, 6off, 7on, 8off

Sure enough, our dip switch settings were identical meaning the disk already
was configured for slip sectoring. It would seem that the disk really is dead
<sniff> and has been retired from service. Ah well.

My thanks to all who replied:

dan@bellcore.com (Daniel Strick)
blazer.babss.basg.com!tkw@bagate.uucp (Terry White)
Geert Jan de Groot <geertj@ica.philips.nl>
strombrg@hydra.acs.uci.edu (Dan Stromberg)
baxter@rocky.mgh.harvard.edu (Larry Baxter)
Tom Slezak <slezak@llnl.gov>
John Valdes <valdes@geosun.uchicago.edu>
peter@key.amdahl.com (Peter Sivo)
glk@pls.amdahl.com (Greg Kastanek)
Birger.Wathne@vest.sdata.no (Birger A. Wathne)
poffen@sj.ate.slb.com (Russ Poffenberger)
bukys@cs.rochester.edu
markw@utig.ig.utexas.edu (Mark Wiederspahn)

-- 
- Mike Jewison, Systems Manager
  Space Astrophysics Laboratory
  Institute for Space and Terrestrial Science
  4850 Keele St., 2nd Floor
  North York, Ontario
  CANADA   M3J 3K1

Internet: mjewison@nereid.sal.ists.ca Phone: (416) 665-5451 FAX: (416) 665-5497



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:58 CDT