0

I have a problem with a software raid running mirror. One disk is dead and the server cannot find it. Normally I would have removed it as well.

mdadm --manage /dev/md0 --remove /dev/sda1

But after it can't see the bad disk I got an error saying it couldn't see /dev/sda1

I then chose to have the disk changed and run the raid again. But if I now do a cat /proc/mdstat it comes up with the following

md3 : active raid1 sda4[2] sdb4[1]

Why is there now a [2] and [1] is it because it still sees the old disk as part of the raid?

If it still sees them as part of the raid, how do I remove the partitions when I can't access them because the disk is dead?

The Tech Guy
  • 21
  • 1
  • 1
  • 3

1 Answers1

0

Your md3 array includes sda4, not sda1.

If sda is gone, you can try to remove it from the array via

mdadm -r /dev/md3 failed
mdadm -r /dev/md3 detached

but be sure it is the correct array (md3).

If in doubt, stop now and post the full output of cat /proc/mdstat and mdadm --detail /dev/md3

shodanshok
  • 47,711
  • 7
  • 111
  • 180