ML370 G3: SCSI Controller Problem?

[Seth Horila]

One of our ML370 G3 machines has had a reoccuring SCSI related issue that I have not been able to solve or HP. The server has a 641 controller with 4 hard drives hooked up to it. I have no problem with the 641 controller or the hard drives connected to them. However there are 2 hard drives in a 2-bay cage connected to the integrated Adaptec controller. These are mirrored in Windows 2003 and are the system drives. In the past I have had reoccuring issues with SMNP traps for temperature status changes for the 2 bay cage. A lot of parts have been replaced in this system, the system board (3x), 2-bay cage, scsi cables, memory and 1 of the 2 processors. Although with software upgrades the SNMP trap errors were fixed a couple months ago. Controller issues started on 11/14 and I have been getting errors:

dmio: Volume Volume2 block 3477362 (mountpoint C:): Uncorrectable write error
dmio: write error on Plex Volume2-01 of volume Volume2 offset 5382517 length 128
dmio: Harddisk0 write error at block 5382580: status 0xc000009d

disk: The device, \Device\Harddisk0, is not ready for access yet.

ADPU160M: "The driver for device \Device\Scsi\adpu160m detected a port timeout due to prolonged inactivity. All associated busses were reset in an effort to clear the condition."
ADPU160M: "The device, \Device\Scsi\adpu160m, did not respond within the timeout period."

I actually received a complete scsi controller failure at boot with the onboard controller. So i hooked up a Adaptec 29160 in place of the onboard and the errors still persist. The onboard contoller shares the same driver as the one I installed in the PCI slot. I am guessing it could be a hard drive related issue. So i broke the mirror, removed disk 0 and put disk 1 in place of disk 0. The problem of the driver adpu160m "The driver for device \Device\Scsi\adpu160m3 detected a port timeout due to prolonged inactivity. All associated busses were reset in an effort to clear the condition." still persists. So I am planning on taking the original disk0 and put that back in to see if it clears the problem. Does anyone know why this problem would persist with two totally different controllers? Could the hard drives be causing this timeout? I have run HP Diagnostics for 10 hours and no errors were detected. I also ran the hddetect util to verify the firmware was okay and that passed.

[sandeep_raman]

Hello Seth,

Couple things to try incase not tried:

1.Check for bad sectors on the disks
2.See if the disk model/part # is listed in

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?lang=en&cc=us&objectID=c00677430&jumpid=reg_R1002_USEN
http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?lang=en&cc=us&objectID=c00792595&jumpid=reg_R1002_USEN

SRH

[Seth Horila]

I have checked the model numbers of both hard drives, they are BF03688284 running firmware HPB2. Even though these model numbers are not affected, would I be able to update the firmware?

[Seth Horila]

I have downloaded the individual and most recent firmware for all the storage disks on this server. I dont know why Firmware Maintenance hasnt been updating the drives in the past. I downloaded the most recent 7.60 Firmware CD. Pre-update the firmwares for the drives are:

On the Integrated SCSI Controller:
BF03688284 HPB2
BF03688284 HPB2

On the 641 Controller:
BD03688272 HPB1
BD036863AC HPB4
BD03688272 HPB4
BD03688272 HPB1

I will update them all tonight and see if that fixes the problem.

[Seth Horila]

I have updated the firmware on all the drives. Come to find out there is a problem with the Firmware Maintenace CD. It does not apply firmware to any HD that is plugged into the integrated system board scsi controller. So you have to pull the HDs out and put them on a Smart controller. All this time, I thought firmware upgrades were being applied to these drives and they were not. If this is the problem, this would have saved HP and myself many-many hours of time troubleshooting. I will let it run a couple days to see if the errors clear. I hope HP fixes this issue.

[Seth Horila]

The firmware updates did not solve the problem. The OS locked up and upon reboot the SCSI Controller failed with 39160 Port 1 at Slot 00, 01:03:001701 Scsi controller failure. I then transferred the HDs over to the Adaptec 29160 controller and it wouldnt scan the drives. I then pulled a hard drive from another server and it detected it. I then shut down the machine and put the original HD back in and it came up like nothing had happened. HP had shipped me a new system board, I am going to replace that and see if that has any impact. Im ready to paint this machine yellow and call it a lemon.