Koozali.org: home of the SME Server

Obsolete Releases => SME Server 7.x => Topic started by: henrikmc on May 28, 2010, 08:49:53 PM

Title: Failed raid
Post by: henrikmc on May 28, 2010, 08:49:53 PM
During last backup, I got a mail from mdadm monitoring: A Fail event has been detected on md device /dev/md2.

its sme raid1

Personalities : [raid1]
md2 : active raid1 hdc2[0] hda2[2](F)
      195253888 blocks [2/1] [U_]

md1 : active raid1 hdc1[0] hda1[1]
      104320 blocks [2/2] [UU]

So device hda partition 2 is taken out of the array - right?

Im then trying to add it back with:
mdadm --add /dev/md2 /dev/hda2

But I then get a:

mdadm: Cannot open /dev/hdc2: Device or resource busy

How should I procede?  :?

 smartctl says "passed" for both hd
Title: Re: Failed raid
Post by: christian on May 29, 2010, 06:52:31 PM
You are almost there. The disk is down but not out.

remove the failed disk partition from the array first with:
Code: [Select]
mdadm /dev/md2 --remove /dev/hda2
and then re-add it in with:
Code: [Select]
mdadm /dev/md2 --add /dev/hda2
I've updated the how-to to address this use case:
http://wiki.contribs.org/Raid#Resynchronising_a_Failed_RAID
Title: Re: Failed raid
Post by: henrikmc on May 29, 2010, 07:31:23 PM
Thx, that did the trick!
Its now rebuilding

Personalities : [raid1]
md2 : active raid1 hda2[2] hdc2[0]
      195253888 blocks [2/1] [U_]
      [>....................]  recovery =  1.7% (3388736/195253888) finish=111.8min speed=28599K/sec
md1 : active raid1 hdc1[0] hda1[1]
      104320 blocks [2/2] [UU]

unused devices: <none>


Henrik