Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 24 Jul 2012 16:30:58 -0400
From:      Michael Powell <nightrecon@hotmail.com>
To:        freebsd-questions@freebsd.org
Subject:   Re: Disk Errors
Message-ID:  <jun0m2$f32$1@dough.gmane.org>
References:  <d65bfc394e4b31d92bb6ab9e8d220d16@dweimer.net> <20120724180421.GF38393@dan.emsphone.com> <e7e1dc818414243c7df4f1c7aca5e834@dweimer.net>

Next in thread | Previous in thread | Raw E-Mail | Index | Archive | Help
dweimer wrote:

[snip]
> 
> SMART Attributes Data Structure revision number: 10
> Vendor Specific SMART Attributes with Thresholds:
> ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
>    1 Raw_Read_Error_Rate     POSR--   117   099   006    -    145191418
[...]
>    7 Seek_Error_Rate         POSR--   078   060   030    -    77590473
[...]
> 195 Hardware_ECC_Recovered  -O-RC-   025   023   000    -    145191418
[...]
> 241 Total_LBAs_Written      ------   100   253   000    -    1480696469
> 242 Total_LBAs_Read         ------   100   253   000    -    922627427
[snip]

Really, most of the numbers don't look really bad, but I'd cast a leery eye 
towards the way these three correlate.  Read errors from bad spots in the 
magnetic media are one thing, but notice how the drive is recovering data 
with built-in ECC routines. Then notice that the seek error rate is moving 
along at a similar pace. There is a possibility that this is a purely 
mechanical weakness in the head positioning function, just barely "not bad" 
enough for to allow the drive to attempt to hide it through ECC.

When I suspect media failure I generally use the manufacturers diagnostic 
utility to scan for defective media. I haven't used many Seagates in a long 
time so mostly this means WD's wddiags, which can be downloaded as a 
bootable CD .iso image. Seagate will have something similar. The quick scan 
is meant to be non-destructive while the long scan usually is. (I just had 
an old Raptor drive grow 5 bad spots recently, and the long scan fixed it 
without destroying any data - a first for me that) 

As long as the remap space area on the drive is not full usually these 
diagnostics have a good chance to fix bad spots. If it's an infrequent affair 
then one  may just continue to use it. If I see new bad sectors a week later 
it is an indication that the drive has outlived it's usefulness and I 
replace it. If it's another year before I get a small handful of bad spots I 
may just let the diags fix it and continue to use. That is - as long as the 
remap space is not full. Once that happens any new bad spots are permanent 
and cannot be done anything about. Time to replace drive.

The difference here is bad spots developing in the media on the platter(s) as 
opposed to the problem actually stemming from head seek position-location 
problems. None of the diags can do anything about head seek troubles, only 
identify if the problem is media on the platter(s) related.

-Mike





Want to link to this message? Use this URL: <http://docs.FreeBSD.org/cgi/mid.cgi?jun0m2$f32$1>