EVA 4000 72.8GB drive failures

[Kevin Spies]

I notice that these fibre drives fail/amber light much faster than say SCSI 320U 73GB drives. Is there a reason for this? I believe the drives show bad after so many soft errors increment but the drive itself is not "bad".

Is there any way to reset the soft error counter or reuse the drives?

Typically the drives fail after 2 years but the SCSI U320 drives I use have been running for 5 years.

[Uwe Zessin]

I could be a bad batch. Some years ago, some customers reported such failures, too, but never asked for model numbers. Some months later I was able to chat a little with one of the engineers responsible for EVA disk drive qualifications, but I said he was not aware of high failure rates. Well, maybe the information was not fed back to him...

Are these disks running with the latest disk drive firmware? I've seen that some models receive updates to their threshold tables from time to time. So the drive physics might be fine, but the software declares it is not.

[Víctor Cespón]

Latest EVA disk firmware bundle here:

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=329290&prodSeriesId=3302267&prodNameId=3664751&swEnvOID=54&swLang=13&mode=2&taskId=135&swItem=co-61570-1

To know the reason for a disk to be marked as failed, you should check the EVA event log. The controllers have limits sets for each kind of event, if the number of events of some kinds on one disk is above the threshold, it is marked as inoperable.
But this can be due to communication problems (cable, I/O module, another disk causing trouble on the loop...).
This requires an detailed analysis of the EVA controller event log.
Also, if a drive has been maked as failed, but it was due to a communication problem and not the drive itself, there's a way to remove it from the DSL and make it work again, but it's an internal HP procedure.

[Kevin Spies]

Thank you for the help! I will check the drive firmware ect.