DL760

[ayates]

Has anyone come across a repeated drive failure problem with these servers? We have one where two drives just failed. The failed again within one week. Then another drive. The common thread is they all failed on the internal drive cage.
The server has crashed, and when it restarts it reports damaged data. One time the disks on the e: drive failed, then it crashed, and reported curruption on c:. It is clearly not a drive problem. Something wrong in the hardware.
We had similar problems with another DL760. Never had this problem on about 70 other DL380 servers, so it seems to be something specific to the way the internal drive cage works on the 760

[Mike Bollman_1]

Anthony we have several 8500 and DL760G2 servers (no G1's though) and have not had an abornmal drive failure rate. Although I have seen this occur before in other servers and it has been a cage issue. If I were you I would make sure to update the firmware on the drives/server/controllers to the latest rev. via the firmware maintenance cd.

http://h18023.www1.hp.com/support/files/server/us/download/22722.html

I would also look into replacing or swapping the cage with another system.

[ayates]

Strange. We have already replaced the cage and the problem got worse: more drive failures - in the same cage, a crash and a failure to start (could not find system disk).
I am wondering if there is an electrical or SCSI fault with the cage connection to the array controller

[e4services]

May be the controller is at fault.

[Terry Hutchings]

Are you using the integrated controller in this server? This could be a midplane problem or possibly system board. If this is using a SA controller in a pci slot, then I would say the likely cause of this problem is the array bypass cable running from the controller and plugging into the system board.

[ayates]

It has a PCI array controller. The arrays in the external drive cage are working fine. The problem is only with the arrays using the internal drive cage. We have replaced the cage and still have the problem. What are the components between the cage and the array controller that could go wrong? Very strange.

[Terry Hutchings]

Not really all that strange. It's likely the either the seating of the cable or the actual cable from the controller to the system I/O board.