SUMMARY: Defective disk problem

From: Demetris Papallis (Demetris.Papallis@cylink.net)
Date: Tue Nov 05 1996 - 04:36:06 CST


Hello,
        Many thanks for some very helpful information go to:

        Thomas White
        Herbert Wengatz
        Jay Morgan
        Brian Desmond

The problem was stated as:
+> Hi,
+> I have a seagate barracuda that's been giving me problems, and I've
+> tried everything to save it from having to format it. The steps I tried were
:
+>
+> Trying to fsck a certain partition I get:
+>
+> root@ipx-3 % fsck /dev/rsd1a
+> ** /dev/rsd1a
+> BAD SUPER BLOCK: TRASHED VALUES IN SUPER BLOCK
+> USE -b OPTION TO FSCK TO SPECIFY LOCATION OF AN ALTERNATE
+> SUPER-BLOCK TO SUPPLY NEEDED INFORMATION; SEE fsck(8).
+>
+> So I resort in doing:
+>
+> root@ipx-3 % fsck -b 32 /dev/rsd1a
+> Alternate super block location: 32
+> ** /dev/rsd1a
+>
+> CANNOT READ: BLK 590960
+> CONTINUE? y
+>
+> THE FOLLOWING SECTORS COULD NOT BE READ: 590960, 590961, 590962, 590963,
+>
+>
+> I then revert to format and attempt to repair the blocks. I do the
+> following procedure for all 4 defective blocks. I even followed up with an
+> analyze (both refresh and test), which shows nothing.
+>
+> format> repair
+> Enter block number of defect: 590960
+> Ready to repair defect, continue? y
+> Repairing block 590960 (497/4/92)...done
+>
+> format> analyze
+> refresh - read then write (doesn't harm data)
+> test - pattern testing (doesn't harm data)
+>
+>
+> However, after all these, when I try fsck again, I get the exact same
,

The answers ranged. The one I found to be more helpful was:

>>>>>>>
Are these the Absolute Block Numbers ??
> THE FOLLOWING SECTORS COULD NOT BE READ: 590960, 590961, 590962, 590963
if not try the Absolute numbers (found in /var/adm/messages)
if they are absolute and this keeps happening...
replace the drive.
You will probably get this trouble again even if you reformat ...

tw
<<<<<<<

Other information were about how intolerant are these particular disks to
heat,and that my disk is screwed and I should just backup everything I can :)

And another very interesting piece of info from Jay, relevant to Solaris and
Barracuda's. Unfortunately (or fortunately :) I run SunOS.

>>>>>>>>>>>>>>>
I'm not sure if this will help, but if you're using a SPARCStorage array,
This may help.

I found out about a patch today that can be crucial to data integrity on
SPARCstorage Arrays. I never knew this was necessary. Those of us who have
worked with the arrays in the past know that the array controller firmware
usually needs to be updated before use, however you may not know about needing
to upgrade the firmware on the disks within the array. The following is an
excerpt from the patch README for 103451.

Jay Morgan

Patch-ID# 103451-02
Keywords: ST32550 2.1GB 0418 0416 ssa sparcstorage array disk firmware
Synopsis: Hardware, Disk: Download program and the firmware 0418 for ST32550 dis
k
Date: Oct/09/96

Solaris Release: 2.3, 2.4, 2.5, 2.5.1

SunOS Release: 5.3, 5.4, 5.5, 5.5.1

Unbundled Product: Hardware/Disk

Unbundled Release: Seagate ST32550WC and ST32550N 2.1GB disks

Relevant Architectures: sparc

BugId's fixed with this patch: 1218177 1224486

Changes incorporated in this version: 1218177 1224486

Patches accumulated and obsoleted by this patch:

Patches which conflict with this patch:

Patches required with this patch:

  SOLARIS 2.3 PATCHES:

        103351-02 or higher "ssd" SparcStorage Array driver
        103479-02 or higher "sd" driver in SSA systems
        101378-18 or higher "sd" driver in non-SSA systems

  SOLARIS 2.4 PATCHES:

        103290-02 or higher "ssd" SparcStorage Array driver
        101945-37 or higher "sd" driver (and 2.4 kernel jumbo patch)

  SOLARIS 2.5 PATCHES:

        103017-04 or higher "ssd" SparcStorage Array driver
        102984-01 or higher "sd"

  SOLARIS 2.5.1 PATCHES:

        103766-02 or higher for SPARCstorage Array

Obsoleted by:
Files included with this patch:
   download, ST32550.0418.
Problem Description:

1218177 Write errors and idle seek implementation
1224486 sd: there should be retries for both read & write in case of media/hw er
ror
**************************************************************************
**************************************************************************
**WARNING:****WARNING:****WARNING:****WARNING:****WARNING:****WARNING:**

1. This package will not function on "non-Sun" ST32550 disks. It will work
   only on Seagate ST32550N and ST32550WC disks having Sun Labels and
   installed in Sun SPARCservers and SPARC desktops.

2. Follow the instructions given below and do not issue any other command
   except the ones specified below. failure to do so could result in
   defective disks. Any disruption during the download process could result
   in defective disks.

3. If the disk to be upgraded is a system disk, then this upgrade is NOT
   recommended. See instructions under General guidelines in Step 3 of
   Patch Installation instructions for possible alternatives.

**************************************************************************
**************************************************************************
Patch Installation Instructions:
--------------------------------

1. SPECIAL NOTE:PLEASE INSTALL THE REQUIRED PATCHES PRIOR TO RUNNING THE
                DOWNLOAD SOFTWARE. ENSURE THAT THE SYSTEM IS FULLY FUNCTIONAL
                (AFTER REBOOTING) WITH THE APPLIED PATCHES BEFORE ATTEMPTING
                TO RUN THE DOWNLOAD SOFTWARE.

2. General guidelines for upgrading:

        All previous firmware versions are upgradable to this level.

        Back up all filesystems.

        Do not upgrade the system disk.

        Stop ALL applications .

        UNMOUNT all file systems on the drives to be upgraded.

        Follow the given instructions during the upgrade and do not
        issue any other command except the ones specified below; do not
        power cycle any drive except as specified below.

        In case of any disruption or unforeseen events happening on the
        relevant scsi bus during the firmware download process, it may be
        that the upgraded disk becomes non-functional. In this
        event, it will be necessary to swap out the disk. This would happen
        as a result of an incomplete or corrupted firmware file
        being downloaded.

**NOTE**If a system disk happens to be a Seagate ST32550 needing a firmware
        upgrade, then one of two options is available. First, the system disk
        could be re-configured so that it is not a system disk at the time of
        firmware download by setting up another disk as a system disk.
        Second, the system disk could be swapped out with a spare disk.

3. Procedure:

The procedure to be used for upgrading the ST32550 drives to 0418 firmware
level is explained below. The time to upgrade is about 1 minute per disk;
the time required the upgreade in SPARC Strorage Array is about 3 minute
per disk.

a) Restore the patch (through uncompress and tar) into any directory ,say,
        <patchdir>

b) cd <patchdir>

c) In the patch directory, as root, type the 'download'
       command as following:

                % su root
                # ./download

d). Enter the appropriate number range for the disks to be upgraded.
          ** See the example below**.

** NOTE** This upgrade will result in error messages in the console
                window and/or the "download" window.

e). When the download is done, the list of the drives will be
    displayed again; the current code of the drives will be
    displayed as 0418. Quit the 'download' program by typing 'q'.

NOTE: WHEN THE DOWNLOAD PROGRAM OR THE FORMAT/INQUIRY COMMANDS INDICATE
THAT THE FIRMWARE LEVELS ARE AT 0418, THEN THE UPGRADE IS SUCCESSFUL
AND COMPLETE IRRESPECTIVE OF THE ERROR MESSAGES THAT MIGHT BE DISPLAYED.

See the patch for more info.

<<<<<<<<<<<<<<<<<<<<

        demetris

Demetris Papallis 10 Katsonis Street,
Inernet Services Administrator Neoelen Marina Building,
Cylink Information Services Office 301-302,
Tel +357-2-317617 Nicosia 1660,
Fax +357-2-316341 Cyprus
http://www.cylink.net P.O.BOX 554



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:11:15 CDT