also sprach harland christofferson <debian-user@audubonstrings.com> [2007.09.19.0405 +0100]:
> okay, this is what i have done:
>
> # umount /dev/hdc8
> # mdadm /dev/md4 --fail /dev/hdc8
> # mdadm /dev/md4 --remove /dev/hdc8
> # mdadm /dev/md4 --add /dev/hdc8
>
> cat-ed /proc/mdstat and see that it is resynching ...
> good news.
>
> now, part two of your suggestion ... unmounting
> /dev/hdaX partitions, editing fstab, and remounting.
>
> won't this cause my system to go belly up?
You need to do this from a rescue disk. I recommend grml.
Even though the above procedure *does* restore your RAID and you
change /etc/fstab such that on next boot, /dev/mdX would be mounted
instead of /dev/hdaX, between the --add and the reboot, something
will write to /dev/hdaX and you're back to where you started from.
If you don't have a rescue disk, here's one way which would probably
also work, but I can't guarantee it.
back up the important files
change /etc/fstab accordingly
fail and remove /dev/hdcX from all arrays
call mdadm --zero-superblock on all /dev/hdcX partitions
reboot
readd all /dev/hdcX partitions to the appropriate arrays
--
.''`. martin f. krafft <madduck@debian.org>
: :' : proud Debian developer, author, administrator, and user
`. `'` http://people.debian.org/~madduck - http://debiansystem.info
`- Debian - when you have better things to do than fixing systems
"and the sea isn't green
and i love the queen
and what exactly is a dream?
and what exactly is a joke?"
-- syd barrett
Attachment:
digital_signature_gpg.asc
Description: Digital signature (see http://martin-krafft.net/gpg/)