Multiple disk error warnings

[snapshot]

I'm getting loads of errors in the messages file about writing to disks but there is no apparent problem with any individual disk. They are all on the same controller c1 Any suggestions as to how I go about determining where the problem lies?




Feb 9 09:16:00 tulip scsi: [ID 107833 kern.notice] #Requested Block: 46 Error Block: 46
Feb 9 09:16:00 tulip scsi: [ID 107833 kern.notice] #Vendor: HITACHI Serial Number: 0348S1NJBN
Feb 9 09:16:00 tulip scsi: [ID 107833 kern.notice] #Sense Key: Unit Attention
Feb 9 09:16:00 tulip scsi: [ID 107833 kern.notice] #ASC: 0x29 (<vendor unique code 0x29>), ASCQ: 0x3, FRU: 0x0
Feb 9 09:56:31 tulip scsi: [ID 107833 kern.warning] WARNING: /pci@1f,700000/scsi@2/sd@2,0 (sd3):
Feb 9 09:56:31 tulip #Error for Command: write Error Level: Retryable
Feb 9 09:56:31 tulip scsi: [ID 107833 kern.notice] #Requested Block: 262932 Error Block: 262932
Feb 9 09:56:31 tulip scsi: [ID 107833 kern.notice] #Vendor: HITACHI Serial Number: 0348S1MLFY
Feb 9 09:56:31 tulip scsi: [ID 107833 kern.notice] #Sense Key: Unit Attention
Feb 9 09:56:31 tulip scsi: [ID 107833 kern.notice] #ASC: 0x29 (<vendor unique code 0x29>), ASCQ: 0x3, FRU: 0x0
Feb 9 10:10:02 tulip scsi: [ID 107833 kern.warning] WARNING: /pci@1f,700000/scsi@2/sd@1,0 (sd2):
Feb 9 10:10:02 tulip #Error for Command: write(10) Error Level: Retryable
Feb 9 10:10:02 tulip scsi: [ID 107833 kern.notice] #Requested Block: 37460902 Error Block: 37460902
Feb 9 10:10:02 tulip scsi: [ID 107833 kern.notice] #Vendor: HITACHI Serial Number: 0348S1NJDQ
Feb 9 10:10:02 tulip scsi: [ID 107833 kern.notice] #Sense Key: Unit Attention
Feb 9 10:10:02 tulip scsi: [ID 107833 kern.notice] #ASC: 0x29 (<vendor unique code 0x29>), ASCQ: 0x3, FRU: 0x0
Feb 9 10:37:13 tulip scsi: [ID 107833 kern.warning] WARNING: /pci@1f,700000/scsi@2/sd@2,0 (sd3):
Feb 9 10:37:13 tulip #Error for Command: write Error Level: Retryable
Feb 9 10:37:13 tulip scsi: [ID 107833 kern.notice] #Requested Block: 262932 Error Block: 262932
Feb 9 10:37:13 tulip scsi: [ID 107833 kern.notice] #Vendor: HITACHI Serial Number: 0348S1MLFY
Feb 9 10:37:13 tulip scsi: [ID 107833 kern.notice] #Sense Key: Unit Attention
Feb 9 10:37:13 tulip scsi: [ID 107833 kern.notice] #ASC: 0x29 (<vendor unique code 0x29>), ASCQ: 0x3, FRU: 0x0
Feb 9 11:28:43 tulip scsi: [ID 107833 kern.warning] WARNING: /pci@1f,700000/scsi@2/sd@3,0 (sd4):
Feb 9 11:28:43 tulip #Error for Command: write Error Level: Retryable
Feb 9 11:28:43 tulip scsi: [ID 107833 kern.notice] #Requested Block: 46 Error Block: 46
Feb 9 11:28:43 tulip scsi: [ID 107833 kern.notice] #Vendor: HITACHI Serial Number: 0348S1NJBN
Feb 9 11:28:43 tulip scsi: [ID 107833 kern.notice] #Sense Key: Unit Attention
Feb 9 11:28:43 tulip scsi: [ID 107833 kern.notice] #ASC: 0x29 (<vendor unique code 0x29>), ASCQ: 0x3, FRU: 0x0
Feb 9 12:21:55 tulip scsi: [ID 107833 kern.warning] WARNING: /pci@1f,700000/scsi@2/sd@1,0 (sd2):
Feb 9 12:21:55 tulip #Error for Command: write(10) Error Level: Retryable
Feb 9 12:21:55 tulip scsi: [ID 107833 kern.notice] #Requested Block: 37460502 Error Block: 37460502
Feb 9 12:21:55 tulip scsi: [ID 107833 kern.notice] #Vendor: HITACHI Serial Number: 0348S1NJDQ
Feb 9 12:21:55 tulip scsi: [ID 107833 kern.notice] #Sense Key: Unit Attention

 

[KenCunningham]

Perhaps it's the controller? Have you tried reseating everything?

 

[daFranze]

can you tell us something about the hardware? What is your host, which cabinet is used to carry the disks, wiring?

Best Regards, Franz
--
Solaris System Manager from Munich, Germany
I used to work for Sun Microsystems Support (EMEA) for 5 years

 

[snapshot]

If only it were that simple. The hardware is 100 miles away on a customer site and down time isn't that easy to come by - I was hopeing to narrow it down remotely if that's possible.

 

[KenCunningham]

Can't someone on site give you the hardware details? Checking that nothing is loose shouldn't be too arduous? Otherwise 100 miles = approx 1½ hours travelling, can you get there to see for yourself?

 

[snapshot]

It sounds like that's going to be my only option! Now where's the car keys ? ........

 

[daFranze]

prtdiag -v" output may help a little bit, at least to see what's the host...

Best Regards, Franz
--
Solaris System Manager from Munich, Germany
I used to work for Sun Microsystems Support (EMEA) for 5 years

 

[KenCunningham]

Agreed daFranz. Snapshot, were you joking about the car keys? If not, how'd it go? Any solution? Whatever, I'd suggest doing an audit of all your remote systems and keeping them to hand to help out in the future!