On 03/10/2018 14:46, Feng Zhang wrote:
While one thing I am also wondering is that if there's any efficient method for small sized storage, say <10 disk? Which can be scary if you want to build a RAID6 with the new 10TB disk(for example, 10X10TB disks). The rebuild time can be very very long.
Don't forget this is "linux raid 10" not standard raid 10. And of course rewriting a 10TB disk will take ages anyway :-)
Bearing in mind also you need a minimum of four disks for raid-6, you'll only need five disks for 60 or 61. Of course, doing 61 over five 10TB disks will chew up a lot of disk space - that's 50TB split into two 25TB arrays, giving you 12.5TB of usable space ...
but it has the advantage that if you do lose a disk, the data needed to recreate it is scattered "randomly" across all the other disks, meaning each surviving drive only gets hammered for 2.5TB of data read to recover the full 10TB.
(Of course with ten 10TB drives, it will only need to read 1.1TB per drive to rebuild a broken drive.)
(Implementing TRIM would also help a rebuild, but that's another "someone will do it if they want it" job :-)
Cheers, Wol
