Failing SCSI disk

It looks like our disk is failing.
How to get the number of relocated blocks ?
Is bad ? 9 uncorrected errors (9 pending relocations ?) isn't that bad
and it didn't change in about 24h.
What does  "failed in segment" mean? there is no LBA of first error
which is odd .

thanks in advance
-- 
Lazy

smartctl version 5.33 [i686-pc-linux-gnu] Copyright (C) 2002-4 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

Device: FUJITSU  MAT3073NP        Version: 0108
Serial number: AAS0P54012FR
Device type: disk
Transport protocol: Parallel SCSI (SPI-4)
Local Time is: Sun Sep  4 02:29:27 2005 CEST
Device supports SMART and is Enabled
Temperature Warning Enabled
SMART Health Status: HARDWARE IMPENDING FAILURE TOO MANY BLOCK
REASSIGNS [asc=5d,ascq=14]

Current Drive Temperature:     36 C
Drive Trip Temperature:        65 C
Manufactured in week 16 of year 2005
Current start stop count:      11 times
Recommended maximum start stop count:  10000 times

http://smartmontools.sourceforge.net/#references

All of the links for 'ATAPI/ATA-7 Specification' are broken.  Just
thought I'd send a heads-up.

This one looks pretty current:

http://www.t13.org/docs2004/d1532v3r4b-ATA-ATAPI-7.pdf

In fact, on the www.t13.org page, they are also showing names that look
like ATAPI-8 now, but I look a look and they still reference Atapi-7.

-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf

Re: weird problems with a Seagate drive

I've talked to Seagate. I'm forwarding their response in hope it will
benefit others with the same problem and doubts.

<quote>

 Hello,

 The Short Test of Seatools is not as accurate as the Extended Test of
 Seatools. We mainly rely on the Extended Test for most-accurate results.

 You may visit our support website for more information on the DST error
 code you are receiveing, but if you are not receiving this error code
 under the Extended Test routine then you have nothing to worry about.

 http://www.seagate.com/support/kb/disc/ref/st_error_codes.html

 DST Error Codes

 00 No Error
 01 DST was Aborted by User
 02 DST was Interrupted by Reset or Power Off
 03 DST Could Not be Run
 04 DST Failed for Reasons Unknown
 05 DST Failed Electrical Element (Read/Write Head Check)
 06 DST Failed Servo Element (Track Following Check)
 07 DST Failed Read Element (A Bad Sector was Found)

 Regards,

Fujitsu power on hours attribute

Hi,

I've got this Fujitsu 2.5" drive.  Apparently, its Power On Hours
attribute is storing the *seconds* not hours of power on time.  Can this
be corrected?  Output is smartctl -a after issuing 'smartctl -s on -o on
-S on -t long'.

smartctl version 5.34 [i686-pc-linux-gnu] Copyright (C) 2002-5 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

=== START OF INFORMATION SECTION ===
Device Model:     FUJITSU MHH2048AT
Serial Number:    01021799
Firmware Version: B818
User Capacity:    4,871,301,120 bytes
Device is:        Not in smartctl database [for details use: -P showall]
ATA Version is:   4
ATA Standard is:  Exact ATA specification draft version not indicated
Local Time is:    Mon Sep  5 15:23:43 2005 CDT
SMART support is: Available - device has SMART capability.
SMART support is: Enabled

=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED
See vendor-specific Attribute list for marginal Attributes.

General SMART Values:
Offline data collection status:  (0x03)	Offline data collection activity
					is in progress.

Re: Fujitsu power on hours attribute

Hi,

Ryan Underwood wrote:

>I've got this Fujitsu 2.5" drive.  Apparently, its Power On Hours
>attribute is storing the *seconds* not hours of power on time.  Can this
>be corrected?  
>

Yes, see section about -v (--vendorattribute) option on smartctl manpage.

Please try:

smartctl -v 9,seconds -a ...

This is only necessary for drives not in the drive database.

Christian

-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf

Re: Fujitsu power on hours attribute

On Tue, Sep 06, 2005 at 06:52:07PM +0200, Christian Franke wrote:
> Hi,
> 
> Ryan Underwood wrote:
> 
> >I've got this Fujitsu 2.5" drive.  Apparently, its Power On Hours
> >attribute is storing the *seconds* not hours of power on time.  Can this
> >be corrected?  
> >
> 
> Yes, see section about -v (--vendorattribute) option on smartctl manpage.
> 
> Please try:
> 
> smartctl -v 9,seconds -a ...

That seems to correct it:
9 Power_On_Seconds        0x0012   001   001   020    Old_age   Always
  FAILING_NOW 17557h+53m+36s

Which corresponds to something like 2 years of power on time.  That does
not get rid of the FAILING_NOW flag though...

> This is only necessary for drives not in the drive database.

Should I forward my original mail to smartmontools-database?

--

-- 
Ryan Underwood, <nemesis <at> icequake.net>

smartctl query without spin-up

Hi,

I'm using noflushd to power-down my hard drives when they aren't being
used, but I still want to log their temperatures. Is it possible to
use smartctl to query the temperature of a hard drive without spinning
it up?

Thanks,

Frederik

--

-- 
http://ofb.net/~frederik/

-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf

Re: Endianess trouble

Some time ago I complained about smartmontools not working for me on
NetBSD/sparc64. I finaly found time to track this down. There are two
issues

 - the os_netbsd.* stuff obviously had never been tested on big endian
   machines and was plain wrong.

 - the NetBSD kernel is "helpfull" and translates the complete response
   page to an IDENTIFY command into host byte order. Smartmontools converted
   it again, giving funny results.

I discussed the second point with the kernel developer in charge of that
subsystem, and we basically agreed that changing that behaviour now is
too late. NetBSD tries to keep binary compatibility, even for exotic ioctls
like the one used here.

The impact is pretty low and the change to smartmontools not very intrusive.
Patches attached - could someone review and add to cvs?

Thanks,

Martin

$NetBSD: patch-ab,v 1.5 2005/09/07 07:32:10 martin Exp $

--- os_netbsd.c.orig	2004-04-20 06:53:26.000000000 +0200
+++ os_netbsd.c	2005-09-06 23:02:45.000000000 +0200
 <at>  <at>  -21,4 +21,5  <at>  <at> 

SATA drives using standard ide driver not working

Hi,

I'm trying to get the following combination working with smartmontools:

     2 x Hitachi HDS722525VLSA80 SATA 250GB
     Highpoint RocketRAID 1640 SATA drive controller
     Gentoo LINUX Kernel Version: 2.6.11-r15

Although the drive controller is a (fake-) RAID controller I'm using  
it as a normal SATA controller and using the standard LINUX drivers/ 
ide driver (drivers/ide/pci/hpt366.c - the controller is reported as  
an HPT374) and software RAID.

The drives are /dev/hde and /dev/hdg.

When I run smartctl:

# smartctl -a /dev/hde
smartctl version 5.33 [i686-pc-linux-gnu] Copyright (C) 2002-4 Bruce  
Allen
Home page is http://smartmontools.sourceforge.net/

Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)

A mandatory SMART command failed: exiting. To continue, add one or  
more '-T permissive' options.

I tried -T verypermissive and specifiying -d ata, but with no  
improvement.

Re: smartd - best practice?

On Tue, 23 Aug 2005, Gary Funck wrote:

> Right now, we run the -H (health test) option via smartd on all our
> system disks, which per the manual page tests for changes in
> pre-fail attributes that signal an imminent disk failure.  We don't
> run self-tests, but I'm thinking that we should.  What is the recommended
> interval for running self tests?  Is there any merit in running the
> short tests, or should we also run the long tests?

On my systems, I run short tests every 24 hours and long tests every 7
days.

Bruce

-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf