SEO

I bought a supposedly new SCSI disk on eBay (Seagate ST3146854LC, 147 GB
15k rpm). According to the format command on Solaris 10, this has 633
primary defects and 18 grown defects. Does the fact there are grown
defects indicate it has been used? I'm very suspicious it is not new at
all.

I'm also more than irritated that it was sent in a very thin jiffy bag,
with no anti-static bag. I'm very tempted to tell the seller he can have
it back, as I've no idea what damage this might have suffered in the
post being in just a jiffy bag. According to the seagate site, they
should be shipped with 2" of foam in a corrugated box. I suspect that it
is a bit OTT, but a silly jiffy bag is a feel taking the mic a bit.




--
Dave (from the UK)

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@althorne.org
Hitting reply will work for a few months only - later set it manually.

http://chessdb.sourceforge.net/ - a Free open-source Chess Database

"Dave (from the UK)" <see-my-signature@southminster-branch-line.org.uk> wrote:
> I bought a supposedly new SCSI disk on eBay (Seagate ST3146854LC, 147 GB
> 15k rpm). According to the format command on Solaris 10, this has 633
> primary defects and 18 grown defects. Does the fact there are grown
> defects indicate it has been used? I'm very suspicious it is not new at
> all.

Any number larger than zero in the grown defect list would me make
suspicious.

Most of the SCSI drives I work with have an empty grown defect list (or
maybe up to 5 entries on just a few drives). If the number is not zero I
monitor the defect list for some time to see if it is still growing.
Sometimes I also run a surface analysis (format -> analyze -> read).

A steadily growing defect list is a good sign for the drive to fail
in the near future.

In the SCSI inquiry the build date is also encoded (four digits WWYY -
(W)eek (Y)ear). So run "format -e" -> scsi -> inquiry to see when the
disk was built. A "new" disk should be no older than 6 months.

Other things to try: get the number of power-on-hours from the drive.
You could try the smartmontools
http://smartmontools.sourceforge.net/
they will also work on SCSI drives on Solaris/SPARC. Maybe you can get
this information from this tool.

--
Daniel

Daniel Rock wrote:
> "Dave (from the UK)" <see-my-signature@southminster-branch-line.org.uk> wrote:
>
>>I bought a supposedly new SCSI disk on eBay (Seagate ST3146854LC, 147 GB
>>15k rpm). According to the format command on Solaris 10, this has 633
>>primary defects and 18 grown defects. Does the fact there are grown
>>defects indicate it has been used? I'm very suspicious it is not new at
>>all.
>
>
> Any number larger than zero in the grown defect list would me make
> suspicious.
>

Yes, it is making me suspicious too. I think I'll send the disk back.

> Most of the SCSI drives I work with have an empty grown defect list (or
> maybe up to 5 entries on just a few drives). If the number is not zero I
> monitor the defect list for some time to see if it is still growing.
> Sometimes I also run a surface analysis (format -> analyze -> read).

I've done a verify on the disk, which did not need to add any more. But
there are still 18 grown errors, which is worrying.
> A steadily growing defect list is a good sign for the drive to fail
> in the near future.

I guess 18 is too many. I've just checked all my other disks, and none
have any grown defects, although one or two keep giving errors, which is
why I was going to replace them.

> In the SCSI inquiry the build date is also encoded (four digits WWYY -
> (W)eek (Y)ear). So run "format -e" -> scsi -> inquiry to see when the
> disk was built. A "new" disk should be no older than 6 months.

I'll look at that.

> Other things to try: get the number of power-on-hours from the drive.
> You could try the smartmontools
> http://smartmontools.sourceforge.net/
> they will also work on SCSI drives on Solaris/SPARC. Maybe you can get
> this information from this tool.
>

Cheers, I'll do that. Any idea what might be normal for a new disk?
Clearly 6360 hours (1 year) would not be reasonable for a new disk, but
would 100, or 1000? I'm not sure how much testing they might get at
Seagate after the counter is set to 0.

--
Dave (from the UK)

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@althorne.org
Hitting reply will work for a few months only - later set it manually.

http://chessdb.sourceforge.net/ - a Free open-source Chess Database

"Dave (from the UK)" <see-my-signature@southminster-branch-line.org.uk> wrote:
> Cheers, I'll do that. Any idea what might be normal for a new disk?
> Clearly 6360 hours (1 year) would not be reasonable for a new disk, but
> would 100, or 1000?

I just checked a few disks. The counter should be close to zero when the disk
arrives at you.

--
Daniel

Daniel Rock wrote:
> "Dave (from the UK)" <see-my-signature@southminster-branch-line.org.uk> wrote:
>
>>Cheers, I'll do that. Any idea what might be normal for a new disk?
>>Clearly 6360 hours (1 year) would not be reasonable for a new disk, but
>>would 100, or 1000?
>
>
> I just checked a few disks. The counter should be close to zero when the disk
> arrives at you.
>

Thank you.

--
Dave (from the UK)

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@althorne.org
Hitting reply will work for a few months only - later set it manually.

http://chessdb.sourceforge.net/ - a Free open-source Chess Database

Daniel Rock wrote:
> "Dave (from the UK)" <see-my-signature@southminster-branch-line.org.uk> wrote:
>
>>Cheers, I'll do that. Any idea what might be normal for a new disk?
>>Clearly 6360 hours (1 year) would not be reasonable for a new disk, but
>>would 100, or 1000?
>
>
> I just checked a few disks. The counter should be close to zero when the disk
> arrives at you.
>


I managed to check this one - as I suspect, it is not need or in good condition:


smartctl version 5.37 [sparc-sun-solaris2.10] Copyright (C) 2002-6 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

Device: SEAGATE ST3146854LC Version: D403
Serial number: 3KN0HQGS
Device type: disk
Transport protocol: Parallel SCSI (SPI-4)
Local Time is: Fri Mar 9 11:22:25 2007 GMT
Device supports SMART and is Enabled
Temperature Warning Enabled
SMART Health Status: FAILURE PREDICTION THRESHOLD EXCEEDED [asc=5d, ascq=0]

Current Drive Temperature: 36 C
Drive Trip Temperature: 68 C
Elements in grown defect list: 18
Vendor (Seagate) cache information
Blocks sent to initiator = 1482912841
Blocks received from initiator = 497223568
Blocks read from cache and sent to initiator = 19484506
Number of read and write commands whose size <= segment size = 9919506
Number of read and write commands whose size > segment size = 38340
Vendor (Seagate/Hitachi) factory information
number of hours powered up = 1131.05
number of minutes until next internal SMART test = 116

Error counter log:
Errors Corrected by Total Correction Gigabytes Total
ECC rereads/ errors algorithm processed uncorrected
fast | delayed rewrites corrected invocations [10^9 bytes] errors
read: 135210 48 67 135325 143957 629.868 43
write: 0 0 0 0 0 292.758 0
verify: 9293750 736 969 9295455 9554127 33331.441 1579

Non-medium error count: 16

SMART Self-test log
Num Test Status segment LifeTime LBA_first_err [SK ASC ASQ]
Description number (hours)
# 1 Background long Completed - 1 - [- - -]
# 2 Background short Completed - 0 - [- - -]

Long (extended) Self Test duration: 2125 seconds [35.4 minutes]

--
Dave (from the UK)

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@althorne.org
Hitting reply will work for a few months only - later set it manually.

http://chessdb.sourceforge.net/ - a Free open-source Chess Database