On 29/12/16 09:56, jdd wrote:
Le 29/12/2016 à 10:48, Wols Lists a écrit :
It is my experience that FULLY HALF and more of the broken arrays that come to the linux raid mailing list for help, are desktop drives and the problem is timeout mismatch.
It only takes one transient error on a desktop drive, to knock it out of the array. Trying to recover causes a second drive to go, and your raid-novice is screaming "the raid ate my data". Desktop drives are fine IF YOU KNOW WHAT YOU'RE DOING. Trouble is, too many people don't...
it's what I understood, but I didn't see how to cope with this (other than changing disks)
what is your experience?
I've got two 3TB Seagate Barracudas, so not only are they desktop drives, they're also the very drive that has a dire reputation (Barracudas on the whole are fine, it was the 3TB ones that were dying left right and centre). That said, I've not had any trouble at all. And with a mirror, trouble with the raid isn't quite so serious as with eg raid-5. There's a script on that web page that alters the linux settings so that you don't have a timeout mismatch. That doesn't stop an apparent hang if you get a problem, though. You MUST run that script EVERY boot. I don't, so if I hit trouble ... it'll be my own fault. That said, I'm planning to rebuild my computer when I can afford it, and I'll probably be buying Seagate NAS drives - dunno why, I much prefer Seagate to WD. Prejudice, I guess. Cheers, Wol -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org