SCSI Errors

[KPS]

Trying to troubleshoot the following errors I'm getting in syslog... Looking for a way to find out which device it is pointing these errors to..

Errors are as follows:

Oct 17 11:41:28 imhqu24 vmunix:
Oct 17 11:41:32 imhqu24 above message repeats 115 times
Oct 17 11:41:28 imhqu24 vmunix: SCSI: Read error -- dev: b 31 0x010000, errno: 1
26, resid: 2048,
Oct 17 11:41:32 imhqu24 above message repeats 115 times
Oct 17 11:41:28 imhqu24 vmunix: blkno: 8, sectno: 16, offset: 8192, bcou
nt: 2048.
Oct 17 11:41:32 imhqu24 above message repeats 115 times
Oct 17 11:41:38 imhqu24 vmunix:
Oct 17 11:41:38 imhqu24 vmunix: SCSI: Read error -- dev: b 31 0x010000, errno: 1
26, resid: 2048,
Oct 17 11:41:38 imhqu24 vmunix: blkno: 8, sectno: 16, offset: 8192, bcou
nt: 2048.

[A. Clay Stephenson · Accepted solution]

Note the dev: b 31 0x010000

The 31 is the major block device number. Do an lsdev and you will find that major block device 31 corresponds to 'disk' (as opposed to 'tape' e.g.).
The 0x010000 is the minor device number.

That will be c1t0d0.

A very simple method is to cd to /dev/dsk and do an ls -l and look for the matching major/minor device numbers tuple.

[melvyn burnard]

You have a bad disk. Decoding it:
dev: b 31 equates to the major number
0x010000 equates to the minor number, or t ehactual disk device.
the quick way to find it is:
ll /dev/dsk |grep 0x010000
This shouild give you the device file. in my example:
ll /dev/dsk|grep 0x01
brw-r----- 1 bin sys 31 0x01f000 Aug 13 11:01 c1t15d0

You can then do lssf on the device file:
lssf /dev/dsk/c1t15d0

This should give you the hardware info, as in:
sdisk card instance 1 SCSI target 15 SCSI LUN 0 section 0 at address 0/0/1/1.15.0 /dev/dsk/c1t15d0
You could also use ioscan -funCdisk to find the device

[Vincent Fleming]

Look in /dev/dsk... do an ls -l there and find an entry that looks like this:

bcrw-r----- 2 root root 31 0x010000 17 Oct 2002 c1t1d1

Note the "31 0x010000" in the middle of the line - that's the major/minor numbers indicated in the first error message. That would be your device.

To isolate exactly which device it is, use ioscan to find the c1t1d1 (or whatever the name is on your system) device, then note the hardware path of the device.

For further information, see the man page for ioscan or the online docs at docs.hp.com.

Good luck!

[Ken Hubnik_2]

Do a lsdev to figure out which dev it is. Then do a diskinfo on that device to see if the disk is faulty

[S.K. Chan]

In addition ..
If it's repeating for 115 times, it's definately time to replace c1t0d0, whatever you got attached to it. If see any other error like .."SCSI reset or SCSI lbolt" it could be due to termination/connection but in your case I think it's purely scsi read problem due to bad sector for example.

[Ashwani Kashyap]

Its a block device with major number 31 and minor number 0x010000 for which the driver is sdisk and class is disk .

So it has to be the disk c1t0d0 .