On Mon, Dec 20, 2010 at 4:52 PM, John Stoffel
<john@stoffel.org> wrote:
>>>>> "Eric" == Eric Martin <eric.joshua.martin@gmail.com> writes:
Eric> Now, the main reason for my email. My mythbox / home file
Eric> server runs linux soft-raid in RAID5 with 3 500GB disks. One
Eric> disk went bad a while ago and the array never rebuilt itself.
So you only had two 500Gb disks left in the array?
Eric> The other day, the second disk went bad. Am I hosed?
Possibly, it depends on how bad the second disk is. What I would do
is try to use dd_rescue to copy the 2nd bad disk onto a new disk
(possibly your original bad disk if you feel brave!) and then try to
re-assemble your raid 5 using that. You might or might not have
corruption in the filesystem, so make sure you run an fsck on it.
Now, in the future, you should run a weekly check of the array for bad
blocks or other problems, so that you get notified if a disk dies
silently. I use the following crontab entry:
#
# cron.d/mdadm -- schedules periodic redundancy checks of MD devices
#
# Copyright © martin f. krafft <madduck@madduck.net>
# distributed under the terms of the Artistic Licence 2.0
#
# By default, run at 00:57 on every Sunday, but do nothing unless the
day of
# the month is less than or equal to 7. Thus, only run on the first
Sunday of
# each month. crontab(5) sucks, unfortunately, in this regard;
therefore this
# hack (see #380425).
57 0 * * 0 root if [ -x /usr/share/mdadm/checkarray ] && [ $(date +\%d) -le 7 ]; then /usr/share/mdadm/checkarray --cron --all --idle --quiet; fi
and I get a nice weekly report on both arrays on my main filerserver
at thome.
Eric> I've been googling for 'rebuild bad linux software raid' but all
Eric> I get is the rebuild command. Also, I don't see any tools that
Eric> will move bad data to another spot on the disk. This is my
Eric> first time using software raid so I'm in a bit over my head.
The first thing is to ask for help on the linux-raid mailing list,
which is hosted on vger.kernel.org.
But somethings you can do to help is to give us more information.
Like:
cat /proc/mdstat
mdadm -E /dev/sd...
or /dev/hd... depending on whether your SATA or IDE drives.
Basically, use the devices you got from the /proc/mdstat output as
your basis.
Give us this output, and we should be able to help you more.
John