I have 8 WD Black with TLER at whatever its at factory (off). I haven't had a disk drop out. Ive done a lot with the array. I think it really depends on the hardware. If they get stuck on a bad block, they'll drop out... But I haven't had that happen yet. Its been 1 year since I made this array and I use it daily, heavily.
If you are really worried about it, go with Seagate... They have dropped the ball in the past with the bricked drives and some such, so I know a lot of people dont trust them anymore... but I've heard good things about their firmware after the 7200.12 bricking debacle. They are also a bit faster at sequential reads and seek time than the WDs, from my testing.
Both Seagate and WD have 5 year warranty so, maybe do a RAID6 so you can have 2 drives fail and then make sure to do Advance RMA so you dont have to run degraded for too long.
So, i'm just (finally) beginning the RAID experience with 2 * 1TB RAID1 (going later to RAID5 or RAID6).
I admit : i'm confused with all the readings i've made about stripe size / ntfs cluster size
Well, is there any simple "thumb rule" for it ?
What i've setup :
RAID1 : ARA / WB Cache / 64k stripe
Disk : i had to opt to GPT i suppose (for later expandings)
NTFS cluster size : 32k/64k ? (main files : big files, divx, iso, etc...)
Thank you very much
P.S. : Sorry to ask it because i know it's a difficult subject but i'm really confused after reading so much posts about adequate stripe/cluster sizes...