I have a Raid 6 configuration with 14 HD (16Tb each one) They are configured in this way:
/dev/md0 /swap
/dev/md1 /boot
/dev/md2 /
/dev/md3 /home
and drives are:
sda,sdb,sdc...sdn
Each unity is configured in this way:
/dev/md0 -> sda1,sdb1,sdc1...sdn1)
/dev/md1 -> sda2,sdb2,sdc2...sdn2)
/dev/md2 -> sda3,sdb3,sdc3...sdn3)
/dev/md3 -> sda4,sdb4,sdc4...sdn4)
Today I found my server down because 2 drives (sda and sdj) failed. I assumed that the raid software would automatically fix the situation excluding the failed drives from the array and notify me with an email. Instead, I've found my server down and I couldn't even restart it, because the raid software kicked out /dev/md2 (which points to the root of the file system).
Since the hosting service documentation don't explain how to manage this situation, but gives only the command for replace,add and remove the drive, my question is: Is it possible that this behavior depends on the fact that I have not chosen "hot swap" type drives during the adding to raid?
Assuming that i wish to use the new drives as 2 hotspare, do i have to use this?
SDA Drive
mdadm /dev/md0 -a --spare /dev/sda1
mdadm /dev/md1 -a --spare /dev/sda2
mdadm /dev/md2 -a --spare /dev/sda3
mdadm /dev/md3 -a --spare /dev/sda4
SDJ Drive
mdadm /dev/md0 -a --spare /dev/sdj1
mdadm /dev/md1 -a --spare /dev/sdj2
mdadm /dev/md2 -a --spare /dev/sdj3
mdadm /dev/md3 -a --spare /dev/sdj4
instead of this?
mdadm /dev/md0 -a /dev/sda1
mdadm /dev/md1 -a /dev/sda2
mdadm /dev/md2 -a /dev/sda3
mdadm /dev/md3 -a /dev/sda4
mdadm /dev/md0 -a /dev/sdj1
mdadm /dev/md1 -a /dev/sdj2
mdadm /dev/md2 -a /dev/sdj3
mdadm /dev/md3 -a /dev/sdj4
Thanks for the help
