5404191 batch failure?
I have a bunch of A5200s running in a database environment.
Over the past week, I've replaced 10 Seagate ST318203FC 18G fiber channel disks that have had 'failed' indications on the front panels of multiple arrays (Either failed, or ports not connected). Now I'm showing two more failed.
All of the above are giving <drive type unknown> in format output and 'error' state in Veritas (or :Disconnected).
I also have several arrays with 5404525 Seagate 36G FC drives, of which I've only had to replace a single drive.
Has anyone else seen an increasing number of failing 18G Cheetah FC disks? I'm suspecting a batch failure, but it seems implausible. I'm skeptical.
Thanks in advance for any reports, be they "Everything's ok here!" or "Yep, mine are dying too"
-K
[822 byte] By [
Foxy] at [2007-11-26 7:16:13]
# 1
From my experience with the A5200, I rate them pretty well bullet proof. In four years, you would be unlucky to replace a couple of drives. There are normally two reasons for the drives failing big time. Wide ranges in humidity and perhaps temperature around the box can cause many failures. Another problem could be the dreaded zinc whiskers.
Are your other arrays in the same location as the 5200's? If not, I would be checking the air con and if you have them on a raised floor, look under it for stuff that should not be there.
The only bad disks that Sun had was the IBM drives. Sun replaced many of them.
Another option is to make sure the disks have the latest firmware on them. Its an old but good reason for massive drive failures.
Are you losing both paths to the same disk?
HTH!!
Stephen
