Array failure on DL760 with 5300 controller

[Nelson_17]

Hello,
We have two DL760 servers which have had in the past several disk problems.
We were in contact with our support partner to solve this problems when one of the servers had one disk problem followed by a second disk problem when the array was beeing rebuilt.

I find this very uncommon: first of all we had in the last 2 years more disk problems with this 2 servers than for all the servers we had in the last 10 years...

Is there any know issue? we are using the latest firmware versions for the server the controller and the disks.

[Uwe Zessin]

Nelson,
it could be a batch of bad disks.

[Nelson Ferreira]

Yes, it could be.
This only happened with 72GB disks.

But this time when both the first and the second disk had a crash, the server was under a high load (and disk access).
The CPU was allways 100%. Could there be any other reason than the physical disks (for instance slow i/o being interpreted as a disk problem when there was in fact none?).

[Uwe Zessin]

A high disk access stresses the devices more and increases the change of a failure. I highly doubt that a high CPU load leads to a false disk alarm. On any half-way competent operation system the I/Os are handled in a privileged mode of the operating system and any user process is preempted.

[Nelson Ferreira]

Is there any know hard drive serial number which could be deffective?
This way we could start other activities (replacing all the drives or so).
We cannot keep the situation as it is waiting for the next time a deffective hard drive (or a problem on the machine itself) occurs...

[Prashant (I am Back)]

Hi,

From the following link.
http://h18007.www1.hp.com/support/files/server/us/index.html

select the server then there is hdd identifier utility that will promt if it is from bad batch. Then we can go for replacement.

Regards
Prashant S.