RAID 5 Single Drive Failure

cmac855

Reputable
Nov 3, 2015
3
0
4,510
HP Proliant DL380 with 3 300 GB HDDs in a RAID 5 configuration running SBS 2011 with AD DS, File/Print server, App server, DHCP/DNS server, and some other minor roles.

I run a small repair shop in a small town and I received a call last week from a business who's server had crashed. After my assessment, I determined that a single drive had failed on their RAID 5 array. Performance had been reduced to an unstable, almost unusable state.

They had no spares, so I had a drive shipped overnight. I replaced the drive and initiated the rebuild, which completed "successfully" with no degradation. I restarted the server and immediately noticed the drives' failure LEDs were flashing consistently. I checked the ORCA messages and saw the controller was reporting a predictive failure on all three drives. I also checked the HP array config utility, which reported the same issue, with hard read errors on all 3 drives.

Unfortunately, I cannot find any information on a similar scenario, so I'm not sure where to go from here. I have done a backup with Windows Server Backup, but the bare metal and system state failed, which I believe is due to using a 3 TB drive. I will be using a bootable imaging media to get a solid backup, but I need some direction to get these errors cleared. I am going to move them to a 6 disk RAID 10 array, if I can get a clean image. If not, it's a rebuild, which will put them down for longer than I'd like to (not to mention the painstaking task of finding all the programs they use and getting solid backups of that data). Anyway, long story short, where do I go from here?
 

cmac855

Reputable
Nov 3, 2015
3
0
4,510


Thanks, JM! Unfortunately, I have already read the recover-raid guide and found no help there. The second link is not a reflection of this scenario - the array is not degraded. It has been rebuilt, my concern is that the all drives are reporting errors following the rebuild, which is either an erroneous report or very bad luck. Because the drives are only 300 GB, I don't believe BOTH of the remaining would have failed during the rebuild, something I would expect with larger drives. The third link is very informative, but doesn't really help outside of being interesting. I do appreciate your assistance, though.
 
it may just be its age and its time has come to pass ??? never said what generation it was ?? the gen5 is now like 6 years old ?? how old are the rest of the drives ?? were they all start up the same date ?? if one fails could mean the rest are right behind ??

are the new drives today and there firmware still fully compatible or got to find them old hard to find ones that are ??

like said here

'' Having a couple of the G3 and the G4 servers, and all I have to add to this is the replacement parts are getting expensive..

Also when a disk broke (and they will.. eventually) we couldn't find the 3,5" disks in stock anywhere, so we had to go for a refurbished disk. It held up for probably 1-2 weeks then it went down again. Probably bad luck, but it could happen to you too!''

don't know I looked but came up with nothing out side of it should of been a simple hot swap

sorry
 

popatim

Titan
Moderator
You may have old firmware that is misreporting the predictive failures. HP usually has a download available for that with windows or linux utilities; plus they usually have a dvd download that will bring all the firmware in the box up to current which is the recommended way to proceed.