Disk failed & replaced; how to rebuild broken logical volume?

JMCraig · 08-20-2009, 05:16 PM

Hi Folks,

Working with Ubuntu 8.10 rescue mode here. A disk went bad and I have replaced it and the software RAID portions are rebuilding from the mirrors. But, in addition to the RAID 1 partitions, the failed disk also had a big chunk that was associated with a logical volume.

I now get this message from lvdisplay:

Couldn't find device with uuid 'IU6FeT-...'
Couldn't find all physical volumes for volume group Extra
(repeats the above twice)
Volume group "Extra" not found

I presume I can just remove it and recreate it--although I'm not sure if the data that was on it will survive that. But, then, I'm not sure what was on it--it sure wasn't anything critical. But, I can't figure out how to reference it even (nothing seems to be in the /dev that looks like what the docs suggest its name should be like /dev/vg00, for instance--there's no vg* anything). So, if I can figure out how to either remove it, or extend it to include the corresponding partition of the new disk, either would do.

Anyone guide me on this one?

Thanks much!

John

TimothyEBaldwin · 08-21-2009, 11:19 AM

The names given in the documentation are only examples, when started it will appear as /dev/Extra

If you physical volume was empty it can be removed using:

Code:

vgreduce --remove-missing Extra

To remove the physical volume discarding any logical volumes partially on it use:

Code:

vgreduce --remove-missing Extra

You can use

Code:

vgchange --partial -a

to start the volume group.

To replace the physical volume see the vgcfgrestore man page - warning this will place parts of the existing contents of the new disk in the files that were on the old disk.