bad disk discovery

[prateek sethi]

Hi Scott,
Thanks for the your quick response.

I have different set of hardware . So that's why I want to know how I can
debug it myself . Is there anyway or procedure using that I can findout
about the situation or the reason for CDB errors or disk command failure?

Right now I am giving detail about the setup where I am getting this issue .

I am using LSI SAS2008 controller and connected with supermicro Enclosure
with freebsd 9.3. 16 different disks are there but only one disk is having
problem. That means contoller and cable are fine.

Faulty disk info are like:-.

*smartctl output is:-*

smartctl -x /dev/da23

=== START OF INFORMATION SECTION ===
Vendor:               SEAGATE
Product:              ST3600057SS
Revision:             000B
Rotation Rate:        15000 rpm
Form Factor:          3.5 inches
Logical Unit id:      0x5000c5007725173f
Serial number:        6SL8YLPC0000N5030DY7
Device type:          disk
Transport protocol:   SAS
Local Time is:        Tue Dec  8 18:20:45 2015 IST
*device is NOT READY (e.g. spun down, busy)*

*Logs:-*

Dec  8 14:12:01 N1 kernel: da23 at mps0 bus 0 scbus0 target 148 lun 0
Dec  8 14:12:01 N1 kernel: da23: <SEAGATE ST3600057SS 000B> Fixed Direct
Access SCSI-5 device
Dec  8 14:12:01 N1 kernel: da23: Serial Number 6SL8YLPC0000N5030DY7
Dec  8 14:12:01 N1 kernel: da23: 600.000MB/s transfers
Dec  8 14:12:01 N1 kernel: da23: Command Queueing enabled
Dec  8 14:12:01 N1 kernel: da23: *Attempt to query device size failed: NOT
READY, Logical unit not ready, cause n*
Dec  8 14:12:01 N1 kernel: ses1: da23,pass26: Element descriptor: 'Slot 24'
Dec  8 14:12:01 N1 kernel: ses1: da23,pass26: SAS Device Slot Element: 1
Phys at Slot 23

*driver versions:-*

dev.mps.0.firmware_version: 15.00.00.00
dev.mps.0.driver_version: 16.00.00.00-fbsd






On Tue, Dec 8, 2015 at 3:15 AM, Scott Long <scott4long at yahoo.com> wrote:

> Hi,
>
> If your situation is accurate and the disk is not responding properly to
> regular
> commands then it’s unlikely that it will respond to SMART commands either.
> Sometimes these situations are caused by a bad cable, bad controller, or
> buggy software/firmware, and only rarely will the standard statistics in
> SMART
> pick up these kinds of errors.  SMART is better at tracking wear rates and
> error rates on the physical media, both HDD and SSD, but even then it’s
> hard
> for it to be accurately predictive or even accurately diagnostic.  For
> your case,
> I recommend that you describe your hardware and software configuration in
> more detail, and look for physical abnormalities in the cabling and
> connections.
> Once that is ruled and and the rest of us know what kind of hardware you’re
> dealing with, we might be able to make better commendations.
>
> Scott
>
> > On Dec 7, 2015, at 11:07 AM, prateek sethi <prateekrootkey at gmail.com>
> wrote:
> >
> > Hi ,
> >
> > Is there any way or tool to find out that a disk which is not responding
> > properly is really bad or not? Sometimes I have seen that there is lot of
> > CDB error for a drive and system reboot makes every thing fine. What can
> be
> > reasons for such kind of scenarios?
> >
> > I know smartctl is the one which can help. I have some couple of question
> > regarding this .
> >
> > 1. What if disk does not support smartctl?
> > 2. How I can do smartest use of smartctl command like which parameters
> can
> > tell that the disk is actually bad?
> > 3. What other test I can perform to make it sure that disk has completely
> > gone?
> >
> >
> > Please tell me correct place to ask this question if I am asking at wrong
> > place.
> > _______________________________________________
> > freebsd-scsi at freebsd.org mailing list
> > https://lists.freebsd.org/mailman/listinfo/freebsd-scsi
> > To unsubscribe, send any mail to "freebsd-scsi-unsubscribe at freebsd.org"
>
>