some background:
I have a volume group that spans on two drives.
one drive has a partition table, and one partion is LVM, and it contain
one lv.
the other has two lv's which overs it all (no partition table).
each lv sits on a single drive.
I started with a single huge lv, but decided to change things to this way
to minimize the damage in an event of a drive failure).
I started to get io errors from the lv on the first disk (the one which
is on the drive with the partition table).
just as I decided to backup the stuff, the lv became unmountable.
I suspected bad blocks in the drive, so I ran badblocks in non
destructable mode on the drive.
it didn't find anything.
I tried to run badblocks on the partition itself (badblocks /dev/hde3),
just to be sure.
still nothing.
this convinced me that the problem is not physical.
I decided to have badblocks access the blocks via lvm, meaning badblocks
/dev/data_vg/data
this found about 40K bad blocks.
I am now in the stage before I attemp to fix the file system (reiserfs)
using the bad blocks list. to recover whatever can be recovered.
do you guys have any better idea, because this will surely cost me some
data - and I am not convinced that
there is nothing else at lvm level that can make things more optimistic.
another thing:
this is a very serious problem, which looks like its originating from LVM.
I would be glad to share whatever info (lvm metadata etc) you guys will
need to identify it, in hope this problem will be eliminated, for the
sake of future victims.
Thanks,
Omry.
_______________________________________________
linux-lvm mailing list
linux-lvm@redhat.com
https://www.redhat.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
