#598576 grub-probe reports "unknown LVM metadata header" for LV on RAID6

Package:
grub-pc
Source:
grub2
Description:
GRand Unified Bootloader, version 2 (PC/BIOS version)
Submitter:
martin f krafft
Date:
2014-10-18 21:21:05 UTC
Severity:
important
#598576#5
Date:
2010-09-30 08:26:01 UTC
From:
To:
Let me know what other info I can provide:

# grub-probe -t abstraction /dev/mapper/julia-root
error: unknown LVM metadata header.
grub-probe: error: cannot find a device for /dev/mapper/julia-root
(is /dev mounted?).

# lvdisplay /dev/mapper/julia-root
  --- Logical volume ---
  LV Name                /dev/julia/root
  VG Name                julia
  LV UUID                xQoCXr-YlLI-hrPP-a1kV-z1o5-ZR2h-y8h2Ar
  LV Write Access        read/write
  LV Status              available
  # open                 1
  LV Size                768.00 MiB
  Current LE             192
  Segments               1
  Allocation             inherit
  Read ahead sectors     auto
  - currently set to     6144
  Block device           253:0

# vgdisplay julia
File descriptor 4 (/dev/urandom) leaked on vgdisplay invocation. Parent PID 17559: bash
  --- Volume group ---
  VG Name               julia
  System ID
  Format                lvm2
  Metadata Areas        1
  Metadata Sequence No  112
  VG Access             read/write
  VG Status             resizable
  MAX LV                0
  Cur LV                6
  Open LV               6
  Max PV                0
  Cur PV                1
  Act PV                1
  VG Size               1.82 TiB
  PE Size               4.00 MiB
  Total PE              476932
  Alloc PE / Size       238466 / 931.51 GiB
  Free  PE / Size       238466 / 931.51 GiB
  VG UUID               kRIsN3-GYlK-sSVq-F08b-GfYe-GIXl-Vltxyw

# pvdisplay /dev/md1
  --- Physical volume ---
  PV Name               /dev/md1
  VG Name               julia
  PV Size               1.82 TiB / not usable 4.00 MiB
  Allocatable           yes
  PE Size               4.00 MiB
  Total PE              476932
  Free PE               238466
  Allocated PE          238466
  PV UUID               BjIf5J-5oLQ-nrb6-zIM3-nyAr-P8lr-rov0nX

# cat /proc/mdstat
Personalities : [raid1] [raid6] [raid5] [raid4]
md1 : active raid6 sdc1[0] sda1[5] sdb1[4] sdd1[1]
      1953517568 blocks super 1.2 level 6, 512k chunk, algorithm 2 [4/4] [UUUU]
      bitmap: 4/8 pages [16KB], 65536KB chunk

unused devices: <none>
#598576#10
Date:
2010-09-30 14:32:10 UTC
From:
To:
[...]

Could I have the output of:

  hd -n65536 /dev/md1

please?  (I'm not quite sure exactly how big the metadata offset tends
to be; hoping it's within the first 64KB.)

Thanks,
#598576#15
Date:
2010-10-02 13:18:30 UTC
From:
To:
also sprach Colin Watson <cjwatson@debian.org> [2010.09.30.1632 +0200]:

Attached.
#598576#20
Date:
2011-01-07 14:36:35 UTC
From:
To:
tag 598576 fixed-upstream
thanks

This bug can be explained by the problem in raid6_recover fixed by Doug
Nazar's patch. As Martin doesn't anymore have the broken setup in
question I mark it as fixed-upstream