Write Back Caching Could Not Be Enabled problem with SC10 Disk Subsystem

[Naga_2]

We have a FC60 connected with 4 SC10 Disk subsystems connected to and HP N4000 server. After one of the disk went bad disk 12 in disk subsystem 2, it was replaced with another HP drive. After the disk is replaced, the LUN was build fine. I am seeing the following output. Can any one has seen this error in FC60s.
I would appreciate your input.
# amlog
::
::
Controller log sense for Subsystem 000800A0B805E310 at Thu Jul 29 13:56:46 2004
Controller Time Stamp = 072904 131743
FRU Code = 0x2C
FRU Code Qualifier = 0x810A
Sense Key = 0x06
Additional Sense Code = 0x3F
Add Sense Code Qual = 0x80
Decoded Field Replaceable Unit Information:
FRU Group = Disk Group
SCSI Channel:ID = 2:12
Decoded SCSI Sense:
The controller set the disk state to "Failed - Write failure"

Controller log sense for Subsystem 000800A0B805E310 at Thu Jul 29 13:56:46 2004
Controller Time Stamp = 072904 131743
FRU Code = 0x00
FRU Code Qualifier = 0x0000
Sense Key = 0x06
Additional Sense Code = 0xA0
Add Sense Code Qual = 0x00
Decoded SCSI Sense:
Write Back Caching Could Not Be Enabled
PrintFile - Previous LRBA is wrong!

[Steven E. Protter]

First impression:

bad disk.

can you use mstm or cstm or xstm to make sure the disk is still good?

diskinfo command?

ioscan

SEP

[Jeff Schussele]

Hi,

Sounds like it's time to use the array display command - amdsp

Use either

amdsp -d

to get info on the disks specifically - or

amdsp -h

to get info on the entire subsystem

Rgds,
Jeff

[Naga_2]

Sorry for the lengthy replay:
This includes some previous history too:
This array is set to HW Raid 5.
Couple of days ago, there were 5 disks failed in disk shelf 2 that have 10 disks.
The HP asked use to reseat them by taking them out, and we did them one after the other with time gap to complete the LUN rebuild. All the disks were OK and the LUNS were rebuilt automatically. Some time later (10 minutes to an hour), 2 disks failed again. Hp asked us to repeat the reseating. We followed and it completed the LUNs and failed again as usual.
Then have replaced 2 disks ( The firmware on these disks is older HP01) in this slots within the last 2 days. The HP CE came on site, and rebooted the Controller and told us that it is not a controller problem. He did not run any diagnostics though. Every time I pull out the disks and put them back the LUNs were built and failed. In the amdsp, it said the disk is optimal. Since this is part of LUN, I can not run the diskinfo command. I was wondering if any one has any problems with having disks with multiple versions in the same enclosure had issues. Sometimes the disk says NO Responsive, some times Write Failed.

When the disk was showing problems (with different disk that was able to build LUN but failed later)

#amdsp -a FC60 when there were issues:

SCSI Channel:ID = 2:12
Enclosure = 1
Slot (0-based) = 9
Disk State = NO RESPONSE
Disk Group and Type = 090036000A3DA951A5 LUN
Capacity = 0.0 GB
Manufacturer and Model = NO_VENDOR NO_MODEL
Serial Number = NO_SERIAL_NUM
Firmware Revision = NO_FW_REV

#amdsp output when there were no issues:

SCSI Channel:ID = 2:12
Enclosure = 1
Slot (0-based) = 9
Disk State = OPTIMAL
Disk Group and Type = 090036000A3DA951A5 LUN
Capacity = 17.0 GB
Manufacturer and Model = HP 18.2G MAP3367NC#HJ
Serial Number = MS012632
Firmware Revision = HPC6