• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Testing a drive in a RAID

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Joined
Dec 13, 2005
I don't know if this has been asked already, but is there a tried and true way to test a single drive in a RAID without pulling the drive or breaking down the RAID?

Over the weekend, I set up a few drives in RAID5 in my desktop, poked at them a little, then copied my Steam folder over and downloaded several other games. Everything good - even kept my case open with a fan blowing at the drive cage just to be on the safe (and overly paranoid) side. Last night the Intel Rapid Storage Windows program started yelling the RAID was degraded, and the RAID manager noted it again on startup this afternoon.

I also have CrystalDiskInfo installed which does appear to be able to access drives in the raid individually, and it shows the suspect drive as good.

Before I go throwing money on another drive, is there anything to download to get a 3rd opinion?

(Already checked the sticky - the link to the WD program is dead)
 

knoober

Member
Joined
Mar 18, 2015
All I can add here is from my own similar experience. Ive got a couple of brand mismatched 1T drives in a striped array through Windows. I was really just testing whether it was as bad as Ive read, but I get a fair read/write boost from it. 1st drive reads 132 MB/s read and 122 MB/s write. 2nd is 111 and 107. When I use Crystal to check after putting in an array, I get 220 read and 212 write.

Here is the part that may be pertinent to you: when I use Crystal to test the disks individually while in the array, I get about 160 and 180 for read speeds (IIRC - I didnt save any screenshots for that particular run) and I *dont* recall write speeds. While this setup is different from yours, you can see that Crystal reads different speeds while the drives are in an array - for me anyway. I hope this helps a little. Good luck. :)
 

Mr.Scott

Beamed Me Up!
Joined
Jun 9, 2013
You cannot. RAID management software isolates the individual drives from direct access. You can only test as a volume.
 
OP
freakdiablo
Joined
Dec 13, 2005
All I can add here is from my own similar experience. Ive got a couple of brand mismatched 1T drives in a striped array through Windows. I was really just testing whether it was as bad as Ive read, but I get a fair read/write boost from it. 1st drive reads 132 MB/s read and 122 MB/s write. 2nd is 111 and 107. When I use Crystal to check after putting in an array, I get 220 read and 212 write.

Here is the part that may be pertinent to you: when I use Crystal to test the disks individually while in the array, I get about 160 and 180 for read speeds (IIRC - I didnt save any screenshots for that particular run) and I *dont* recall write speeds. While this setup is different from yours, you can see that Crystal reads different speeds while the drives are in an array - for me anyway. I hope this helps a little. Good luck. :)

Can CrystalDiskInfo run a benchmark though? CrystalDiskMark can only hit the RAID, not the disk itself...

You cannot. RAID management software isolates the individual drives from direct access. You can only test as a volume.

Huh, weird that CDI sees individual drive info then. Well I just remembered these things have a 5 year warranty and got them new back in 2014. I'll send WD an email see what they say - all I have on the array are steam games and some *nix ISOs/VMs. if I lose the array to another failure, so be it.
 

knoober

Member
Joined
Mar 18, 2015
Can CrystalDiskInfo run a benchmark though? CrystalDiskMark can only hit the RAID, not the disk itself...



Huh, weird that CDI sees individual drive info then. Well I just remembered these things have a 5 year warranty and got them new back in 2014. I'll send WD an email see what they say - all I have on the array are steam games and some *nix ISOs/VMs. if I lose the array to another failure, so be it.
Well... I ran the bench and got results, but question how valid they are. Mr. Scott's comment leads me to believe that they are worthless results. YMMV