How to identify a bad disk from disk array box ?

[Deepu Chakravarty]

I have one disk bad. Which /etc/lvmtab tells is c0t5d0 and ioscan says it is "
disk 0 10/0.5.0 sdisk CLAIMED DEVICE SEAGATE ST39175LC
/dev/dsk/c0t5d0 /dev/rdsk/c0t5d0"

My problem is to identify this disk in the disk array. The disk does not say any no. like SEAGATE ST39175LC but model no. and serial No.
Pls help to find out the bad disk.

[Dave Olker]

Hi Deepu,

What kind of array is this disk installed in? I have a couple of VA arrays (VA7110 and VA7410) and both of them support the StorageWorks CommandView SDM tool, which allows me to see the arrays in great detail.

A couple of months ago one of the disks in my VA7410 went bad and I was able to look at the array with my web browser and CommandView SDM and immediately identify which disk was bad. CommandView showed a graphical representation of the array and showed which disk was bad.

Also, when I looked at the array itself, the disk that went bad had a different LED on (yellow/orange) instead of the green LED, which again showed me which disk to replace.

I don't know if your array supports a graphical tool like CommandView SDM, but perhaps the LEDs on the array itself will give you a clue as to which disk is bad.

Regards,

Dave

[Robert-Jan Goossens]

Hi,

Maybe there is a faster methode, but I use the dd command to identify a bad disk in a array.

# dd if=dev/rdsk/cxtydz of=/dev/null bs=512k

Now look at the leds on the array.

Hope this helps,
Robert-Jan

[Deepu Chakravarty]

Hi,
Server is K-370.
Array is:
HA Deskside Storage Enclosure.
Product no:A3311A, serial No.USM2024349.

[Shaikh Imran]

Hi,
What type of array you have ?
Anyways just a try ..
Try amdsp command
( see man amdsp for more details)
If you are lucky you will get
the disk location in the array with some channel ID

Regards,

[Sยภเl Kย๓คг]

Deepu
Pls provide the model no of the Disk Array.Your Disk is Seagate ST39175LC Barracuda 18LP 9.1GB Ultra SCSI-II 7200RPM .
However you can physically observe in the array.If a disk failed normally the LED indicator on the specified disk will give a different colour,yello/orrange.
regards
SK

[Sยภเl Kย๓คг]

pls provide your email address which supports >1.4 MB attachment.I can sent you the service manual of your array
regards
SK

[Deepu Chakravarty]

SK,
My email is amitr@asahiindia.com

[Deepu Chakravarty]

Reply to Robert-Jan:
dd -if command gives following error:

dd read error: Invalid argument.

[melvyn burnard]

You do not have a disk array, this is a JBOD storage enclosure.
At the back willl be the SCSI address settings, and you need to look for one that has it's address set to 5.
You could also try doing
dd if=/dev/rdsk/c0t5d0 of=/dev/null and see which disk light comes on and stays on.

[Deepu Chakravarty]

My disk which gives problem is c0t5d0.
so my dd -if statment is dd -if=/dev/rdsk/c0t5d0 of=/dev/null, gives error
dd read error: Invalid argument

[Tom Danzig]

You have a - in front of the if staement. USe the command below:

dd if=/dev/rdsk/c0t5d0 of=/dev/null bs=32k count=100

This should light up the disk for about 5-10 seconds or so. Increase the count to make it stay on longer.

[Deepu Chakravarty]

Pls see the following command and its output:

# dd if=/dev/dsk/c0t5d0 of=/dev/null bs=32k count=100
/dev/dsk/c0t5d0: Device busy
dd: cannot open /dev/dsk/c0t5d0

# dd if=/dev/rdsk/c0t5d0 of=/dev/null bs=32k count=100
dd read error: Invalid argument
0+0 records in
0+0 records out

Which is the right command ? Both the command does not bring any lights on the disk.

[Tom Danzig]

The correct command used the rdsk device file. You should get output something like this:

# dd if=/dev/rdsk/c0t5d0 of=/dev/null bs=32k count=100
100+0 records in
100+0 records out

Try the command using the device path for one of the other disks in the array (dev/rdsk/c0t4d0 perhaps?). Maybe the disk is totally dead?

[Deepu Chakravarty]

Yes, it looks like the disk is totally dead. But try 'dd' on the other right disk, would not make that disk corrupt ?