On Mon, Jun 10, 2019 at 11:22 PM David Teigland <teigland@redhat.com> wrote:
On Mon, Jun 10, 2019 at 10:59:43PM +0300, Nir Soffer wrote:
> > [root@uk1-ion-ovm-18  pvscan
> >   /dev/mapper/36000d310056978000000000000000014: Checksum error at offset
> > 4397954425856
> >   Couldn't read volume group metadata from
> > /dev/mapper/36000d310056978000000000000000014.
> >   Metadata location on /dev/mapper/36000d310056978000000000000000014 at
> > 4397954425856 has invalid summary for VG.
> >   Failed to read metadata summary from
> > /dev/mapper/36000d310056978000000000000000014
> >   Failed to scan VG from /dev/mapper/36000d310056978000000000000000014
> This looks like corrupted vg metadata.

Yes, the second metadata area, at the end of the device is corrupted; the
first metadata area is probably ok.  That version of lvm is not able to
continue by just using the one good copy. 

Can we copy the first metadata area into the second metadata area?

Last week I pushed out major changes to LVM upstream to be able to handle
and repair most of these cases.  So, one option is to build lvm from the
upstream master branch, and check if that can read and repair this

This sound pretty risky for production.

> David, we keep 2 metadata copies on the first PV. Can we use one of the
> copies on the PV to restore the metadata to the least good state?

pvcreate with --restorefile and --uuid, and with the right backup metadata

What would be the right backup metadata?
could probably correct things, but experiment with some temporary PVs

Aminur, can you copy and compress the metadata areas, and shared them somewhere?

To copy the first metadata area, use:

dd if=/dev/mapper/360014058ccaab4857eb40f393aaf0351 of=md1 bs=128M count=1 skip=4096 iflag=skip_bytes

To copy the second metadata area, you need to know the size of the PV. On my setup with 100G
PV, I have 800 extents (128M each), and this works:

dd if=/dev/mapper/360014058ccaab4857eb40f393aaf0351 of=md2 bs=128M count=1 skip=799

gzip md1 md2