error messages

[Louis McCurry]

a few days ago, the following messages started appearing in the /var/adm/messages file. The last four lines are repetative through today.
The system configuration, as far as disks are concerned is four physical 9GB disks in two volume groups (vg0 and vg1), mirrored. What is the key to breaking down the message to determine the erroring device.

Thanks.

Oct 24 07:25
scb->cdb: 2a 00 00 36 a5 50 00 00 10 00
SCSI: Abort abandoned -- lbolt: 69314221, dev: 1f005000, io_id: a85511, status:
200

SCSI: Async write error -- dev: b 31 0x005000, errno: 126, resid: 8192,
blkno: 992504, sectno: 1985008, offset: 1016324096, bcount: 8192.

SCSI: Read error -- dev: b 31 0x005000, errno: 126, resid: 2048,
blkno: 8, sectno: 16, offset: 8192, bcount: 2048.
LVM: vg[0]: pvnum=0 (dev_t=0x1f005000) is POWERFAILED

SCSI: Write error -- dev: b 31 0x005000, errno: 126, resid: 10240,
blkno: 2310, sectno: 4620, offset: 2365440, bcount: 10240.

SCSI: Read error -- dev: b 31 0x005000, errno: 126, resid: 2048,
blkno: 8, sectno: 16, offset: 8192, bcount: 2048.

[nancy rippey · Accepted solution]

I looks like the device with the problem is
b 31 0x005000
Look under /dev/dsk and grep for 0x0050000.
It will give you the device number
Hope this help.

nancy

[James R. Ferguson]

Hi Louis:

Knowledge Base document #KBRC00000668 provides the answer. The disk device can be determined by the using the dev_t value. For example:

dev_t value of 0x1c045000 is associated with c4t5d0

04 = instance number
5 = SCSI address number
0 = LUN number

Regards!

...JRF...

[Eileen Millen]

The easiest way is to do an ll of /dev/dsk
and match the last 4 or 5 digits.
This is the explanation:
Within the syslog.log file and the dmesg output there can be an entry when scsi errors are occurring. The complaints could be pertaining to SCSI Bus Resets, Parity Errors, Lbolt errors. Along with these errors will come a hex device address. An example of this address would be 0x1f000100 and can be referred to as the dev_t number.

The following is the information needed to understand the translation:

- the 0x signifies a hex number
- the first 2 digits represent the major number in hex - convert these
2 digits to a decimal number
- the remaining digits represent the minor number in hex
- then perform an ll on the /dev/dsk directory to match the major and
minor numbers to an actual device

The following is an example of the process:

eg. 0x1f000100
1f = 31 decimal (which is the major number) 000100 already in hex
(which is the minor number)

ll /dev/dsk, you should see a device file with a major number of 31
and minor number of 000100, this is the device complaining in syslog
or dmesg etc.

Eileen

[Robin Wakefield]

Hi Louis,

If you "ll /dev/dsk" and look for

major=31
minor=0x005000

that will identify the disk in question. It looks like c0t5d0.

Rgds, Robin.