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

Ok, so I'm stupid and made a serious and grave mistake. On my linux box running Ubuntu 10.10, I have a 1.5TB RAID array that I had originally setup using the instructions located here: The RAID array is a 4 x 500GB RAID 5 array with an ext3 formatted filesystem and is not my boot device. The RAID array was just used for storage. The LVM setup is a single physical volume (the RAID device) and a single logical volume that makes up the entire physical volume.

So here is the list of steps that I ended up doing, based on the contents of the files in /etc/lvm/archive (one file was created before each step, and the file lists what the command was):

lvremove /dev/raid_vg/lv0
lvcreate -l 357702 raid_vg -n lv0
lvremove raid_vg
lvcreate -l 357702 raid_vg -n lv0
lvremove raid_vg
vgexport raid_vg
vgimport raid_vg
lvremove raid_vg
vgcreate raid_vg /dev/md1

So, I accidentally called lvremove instead of lvchange -a n Then I realized my mistake and tried to restore the logical volume. But that made the disk seem empty.

Now, I'm not sure, but I don't think the data was wiped (at least not all of it). I've cloned the device with ddrescue just so I don't loose anything and can experiment on the image.

Now, my question is then, is it possible to restore the LVM information to the disk and get my filesystem back? I've got the configuration files in /etc/lvm/archive which give me the start extents, sizes and extent counts.

Any help that anyone might be able to provide would be extremely appreciated. I was an idiot and screwed myself. I just hope that it was temporary and I can get my data back.

share|improve this question
up vote 3 down vote accepted

There is a useful post on LinuxQuestions by rayfordj that suggest that it may be possible to restore a vg/lv, even after a sequence of lvremove && vgremove && pvremove.

The sequence on the posting is:

  1. pvcreate --restorefile /etc/lvm/archive/ --uuid dev/sdX
  2. vgcfgrestore -f /etc/lvm/archive/ --verbose thisVG
  3. vgdisplay -v /dev/thisVG && vgchange -ay /dev/thisVG
  4. mount /dev/thisVG/missingLV /mnt

(Please check these steps for your situation carefully before using, or use the text from the original posting which is much clearer!).

Step number 4. will probably fail and will require a file system check.

Following your description of the problem you may be able to jump to step 2.

There is also an article on LinuxJournal about recovering data from a failed RAID array and LVM vg here.

Considering these two sources, it seems to me that you should be able to diff the current lvm config file and the archived one, and try using the old one instead of the new one. If you are lucky restarting lvm will magically allow you to see the lv with lvdisplay, and then mount it.

share|improve this answer
I used a similar procedure awhile ago and can confirm that pvcreate --restorefile ... approach works. – dtoubelis Apr 29 '11 at 3:38

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.