Originally posted by AtW
View Post
-
Why did it crash then?Vote Corbyn ! Save this country ! -
That's a million dollar question.Originally posted by fullyautomatix View Post
Either way we've learnt a few things today, good thing I never trusted that capitalist RAID thingy ...Comment
-
There is no such thing as 100% reliability.Originally posted by fullyautomatix View Post
A proper RAID adapter will be doing patrol reads in order to catch bit rot, which means that it will spot errors that might otherwise go unnoticed (kind of the point really). When you have a unrecoverable read error rate of 1 in every 10^15 bits read then on an array using 6 x 2TB disks you would expect to see an unrecoverable read error every 10th time through or so, on average. Obviously you need to factor in reads from actual workload as well. So the chances of two errors on two separate disks aren't as slim as you might imagine.
Also, drives fail more than the MTBF numbers suggest and are correlated so after a single failure, another is actually more likely.
http://labs.google.com/papers/disk_failures.pdf
http://www.usenix.org/events/fast07/.../schroeder.pdfWhile you're waiting, read the free novel we sent you. It's a Spanish story about a guy named 'Manual.'Comment
-
Hmm, some directories on that disk got corrupted - so much for "rebuild" RAID6.
Time to format and copy data from backup.
Comment
Comment