Server 08 Software RAID 5 "At Risk"

C

CiscoInside

Gawd
Joined
Mar 26, 2010
Messages
546
I know Software RAIDs are not the best, but at the time it was the only option I had. Rather have some form of RAID then none.

Anyways, I have a RAID 5 setup with 4 identical WD 500 gig Blue drives. This setup has been running for a couple of months fine. I recently checked under disk management to find then 1 of my drives had a status of "Errors". I reactivated the drive, and all the drives re-synced.

Now I am showing two drives out of the 4s with warnings. A different drive now saying "Errors" and the other drive that did show Errors now say online but still has a warning. All drives are now stating "At Risk".

Problem is this program is very lacking of information. All I know is Failed Redundancy, but thats it. Is there anything out there that can check the drives to see what actually is going on.

My only other step would be pulling each drive out and running a spinrite on it to see if the drive is physically failing.

I am backing up everything just incase, it might be the time to spend some money on a real card. But for now, any suggestions?
 
Software RAID generally is the best, but not on Windows platform.

You may have bad sectors on your disks, check the SMART data with an application that can, try HDTune for example, or otherwise boot up a Linux livecd to circumvent Windows. Check for SMART-value "Current Pending Sector", preferably post some screenshots over here so we can look at it.

You can fix the bad sectors with various utilities, Spinrite comes to mind, but it's not free.
 
Thanks I will try that. Well if it does come to bad sectors I have Spinrite at work, so thats not a biggie. But i will post the info over here.

Thanks!
 
Your first mistake was re-activating a drive without checking it first. It's generally never a good idea to ignore when a RAID setup tells you it doesn't like a drive. Pull it out and get it connected to something that will allow you to check the drive more thoroughly.

Most drive vendors have their own diagnostic program. Most expect the drive to be physically connected to a 'standard' drive interface (as in, not behind a hardware RAID setup, be that motherboard or add-in card).

Hardware RAID generally tends to have easier setup, better error handling, faster performance and lower CPU requirements. Software RAID is free.

If you're at the point where two of the four drives are questionable it becomes worthwhile considering four new drives. Set them up in a RAID10, you'll get better performance and reliability (vs RAID5).
 
Well if the drives are going, then I will warranty them out. They are only a few months old, but if that is the case I am more now also looking at a hardware RAID as well since I might have to do a rebuild.
 
sub.mesa said:
You may have bad sectors on your disks, check the SMART data with an application that can, try HDTune for example, or otherwise boot up a Linux livecd to circumvent Windows. Check for SMART-value "Current Pending Sector", preferably post some screenshots over here so we can look at it.
Click to expand...

+1

Definitely check it if you can with Ubuntu Live CD. Screenies would be good too.

If I were you OP, I would definitely send the drives back if they are under warranty, Blue drives should last much longer than that, so you should be good with new ones.
 
Last edited:
Another thought, if you have a high proportion of drives failing, there may be another cause.
 
You must log in or register to reply here.
Back
Top