IO errors

[andrew dutton_1]

Hello all... I am getting some disk errors... or I should say IO subsystem. I think they might be an issue with not have the correct drivers, or a patch might be needed... or who knows I might have a bad drive or bus. Anything anyone can add if greatly appreciated. thanks

HP 11.11 N4000

May 9 09:13:25 isqpro03 vmunix: SCSI: First party detected bus hang -- lbolt: 50114720, bus: 3
May 9 09:13:25 isqpro03 vmunix: lbp->state: 5060
May 9 09:13:25 isqpro03 vmunix: lbp->offset: f0
May 9 09:13:25 isqpro03 vmunix: lbp->uPhysScript: 85fef000
May 9 09:13:25 isqpro03 vmunix: From most recent interrupt:
May 9 09:13:25 isqpro03 vmunix: ISTAT: 01, SIST0: 00, SIST1: 00, DSTAT: 84, DSPS: 00000010
May 9 09:13:25 isqpro03 vmunix: lsp: 0000000000000000
May 9 09:13:25 isqpro03 vmunix: lbp->owner: 0000000059ff8d00
May 9 09:13:25 isqpro03 vmunix: bp->b_dev: cb03b000
May 9 09:13:25 isqpro03 vmunix: scb->io_id: 33a39ba
May 9 09:13:25 isqpro03 vmunix: scb->cdb: 1b 00 00 00 01 00
May 9 09:13:25 isqpro03 vmunix: lbolt_at_timeout: 50112520, lbolt_at_start: 50112520
May 9 09:13:25 isqpro03 vmunix: lsp->state: 10d
May 9 09:13:25 isqpro03 vmunix: scratch_lsp: 0000000059ff8d00
May 9 09:13:25 isqpro03 vmunix: Pre-DSP script dump [ffffffff85fef030]:
May 9 09:13:25 isqpro03 vmunix: 78346000 0000000a 78351800 00000000
May 9 09:13:25 isqpro03 vmunix: 0e000002 85fef540 e0100004 85fef790
May 9 09:13:25 isqpro03 vmunix: Script dump [ffffffff85fef050]:
May 9 09:13:25 isqpro03 vmunix: 870b0000 85fef2d8 98080000 00000005
May 9 09:13:25 isqpro03 vmunix: 721a0000 00000000 98080000 00000001
May 9 09:13:26 isqpro03 vmunix: SCSI: Resetting SCSI -- lbolt: 50114820, bus: 3
May 9 09:13:26 isqpro03 vmunix: SCSI: Reset detected -- lbolt: 50114820, bus: 3
isqpro03: (11i INTER)#

[Kent Ostby · Accepted solution]

Here is the key line:

bp->b_dev: cb03b000

Do an:

ll /dev/dsk | grep 03b000

If something shows up, it's probably that disk.

If nothing does, do: ll /dev/rmt | grep 3b000

If it shows up in /dev/rmt, it's a tape drive.

I would check the termination on the bus and/or open a hardware case to have the device checked out.

[Steven E. Protter]

Shalom

dmesg -

Do some disk i/o covering all disks

dmesg

Or run cstm/mstm/xstm and exercize all disks, even SAN lcoated LUNS.

If you have recently switched out a hotswap disk the first command will clear the error.

Otherwise as noted you have a disk issue as noted in the first response.

SEP

[Jeff Schussele]

Hi Andrew,

Kent's correct in the key line in there.
But I also think the key in the line is cb - that should be the major number of the device in hex - 59 in decimal.
So try the following command:

find /dev -exec ll {} \; | grep 59

If you find a 59 in the major number column (column 5) that will be the culprit - or be the same type device. Then decode the 03b000 - for instance if this was a disk that would be c3t11d0.

HTH,
Jeff

[andrew dutton_1]

Yes it looks like you all were right...

isqpro03: (11i INTER)# ll /dev/dsk | grep 03b000
brw-r----- 1 bin sys 31 0x03b000 Feb 9 14:26 c3t11d0


So it looks like c3t11d0 is the culprit and needs replaced then. Now I have to figure out how to move it or remove it and get it replaced but I should be able to do that through lvm and sam I just don't remember all the commands. Thanks for all the help, i was hopeing it was not a HW issue though.