Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

3 years ago a co-worker setup a software RAID-6 array on Ubuntu 9.04 and I'm getting messages from the OS that the drive has bad sectors and should be replaced. I'd like to remove this drive and replace it with a new drive, however, I have never done this before and I'm terrified that in the process of fixing the array I'm going to end up ruining it.

I know the device ID of the array and I know the device IDs of the individual drives in the array. Additionally I physically have the bad drive.

What are the steps to replace the bad drive with a new drive and get the array running again?

share|improve this question
Sounds like you could do with getting a pro in to be honest, don't mean to be rude but you don't sound too comfortable with this and it's not worth potentially blowing your system out over. – Chopper3 Sep 14 '12 at 14:40
up vote 5 down vote accepted

I strongly suggest you become familiar with the tools and implications of each step before proceeding. The last step has the most potential for error and makes the assumption your boot device is hd0. Know thy bootloader!

In this example sdc is the failing device raid member, sdb is a known good device raid member, and md0 is the raid device.

  1. Create a partition table structure backup of the current (failing) drive:

    # sfdisk -d /dev/sdc > ~/raid.ptbackup

    If this does not work, or the table is not-recoverable/suspect, copy the structure from another drive in the raid set.

    # sfdisk -d /dev/sdb > ~/raid.ptbackup
  2. Mark all failing drive partitions as failed and remove. cat /proc/mdstat for list of raid set partitions and repeat commands for each partition. (sdc1, sdc2, sdc3, etc.)

    # mdadm --manage /dev/md0 --fail /dev/sdc1
    # mdadm --manage /dev/md0 --remove /dev/sdc1
  3. Physically replace failed drive with new drive.

  4. Partition the new drive with the required partition table structure. confirm device name sdc, may have changed when drive replaced

    # sfdisk /dev/sdc < ~/raid.ptbackup
  5. Add drive partitions into raid set. repeat for all partitions as in step 2.

    # mdadm --manage /dev/md0 --add /dev/sdc1
  6. Install grub on new drive.

    grub> device (hd0) /dev/sdc
    grub> root (hd0,0)
    grub> setup (hd0)
    grub> quit
share|improve this answer
If the array is only used for storage, do I still need to install grub on the new drive?. (I have a separate drive with / and /home/ partitions.) – slayton Sep 14 '12 at 17:06
@slayton If it's not bootable, then you can (and should) skip the step entirely. – h0tw1r3 Sep 14 '12 at 17:12
Thank you so much, these instructions were perfect! – slayton Sep 14 '12 at 21:21

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.