[linux-lvm] recovery of a VG - how?

[linux-lvm] recovery of a VG - how?

[lejeczek]

[-- Attachment #1: Type: text/plain, Size: 3622 bytes --]

hi guys

I have a VG which I try to restore:

$ vgcfgrestore --file /etc/lvm/archive/chenbro0.0_00308-2121179640.vg
chenbro0.0
  /dev/mapper/chenbro_a: Checksum error at offset 428544
  Couldn't read volume group metadata from /dev/mapper/chenbro_a.
  Metadata location on /dev/mapper/chenbro_a at 428544 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_a
  Failed to scan VG from /dev/mapper/chenbro_a
  /dev/mapper/chenbro_e: Checksum error at offset 422912
  Couldn't read volume group metadata from /dev/mapper/chenbro_e.
  Metadata location on /dev/mapper/chenbro_e at 422912 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_e
  Failed to scan VG from /dev/mapper/chenbro_e
  /dev/mapper/chenbro_m: Checksum error at offset 422912
  Couldn't read volume group metadata from /dev/mapper/chenbro_m.
  Metadata location on /dev/mapper/chenbro_m at 422912 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_m
  Failed to scan VG from /dev/mapper/chenbro_m
  Scan of VG chenbro0.0 from /dev/mapper/chenbro_p found mda_checksum
a6dbb1e8 mda_size 5723 vs previous c6a5d174 5727
  /dev/mapper/chenbro_s: Checksum error at offset 422912
  Couldn't read volume group metadata from /dev/mapper/chenbro_s.
  Metadata location on /dev/mapper/chenbro_s at 422912 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_s
  Failed to scan VG from /dev/mapper/chenbro_s
  /dev/mapper/chenbro_v: Checksum error at offset 422912
  Couldn't read volume group metadata from /dev/mapper/chenbro_v.
  Metadata location on /dev/mapper/chenbro_v at 422912 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_v
  Failed to scan VG from /dev/mapper/chenbro_v
  Scan of VG chenbro0.0 from /dev/mapper/chenbro_w found mda_checksum
a6dbb1e8 mda_size 5723 vs previous c6a5d174 5727
  /dev/mapper/chenbro_y: Checksum error at offset 422912
  Couldn't read volume group metadata from /dev/mapper/chenbro_y.
  Metadata location on /dev/mapper/chenbro_y at 422912 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_y
  Failed to scan VG from /dev/mapper/chenbro_y
  Scan of VG chenbro0.0 from /dev/mapper/chenbro_z found mda_checksum
a6dbb1e8 mda_size 5723 vs previous c6a5d174 5727
  Scan of VG chenbro0.0 from /dev/mapper/chenbro_aa found mda_checksum
a6dbb1e8 mda_size 5723 vs previous c6a5d174 5727
  /dev/mapper/chenbro_ab: Checksum error at offset 422912
  Couldn't read volume group metadata from /dev/mapper/chenbro_ab.
  Metadata location on /dev/mapper/chenbro_ab at 422912 has invalid
summary for VG.
  Failed to read metadata summary from /dev/mapper/chenbro_ab
  Failed to scan VG from /dev/mapper/chenbro_ab
  Couldn't find device with uuid 3UQeFN-hOMe-nwi9-Ld1D-0bNH-JSvO-EaPclH.
  Couldn't find device with uuid qcmpwi-AAbV-lpl3-3B0U-2aeS-VzJ6-ZEMNYo.
  Couldn't find device with uuid DpJTbc-t0Oe-BImz-pdlo-TkQz-G0Hu-ojSekc.
  Couldn't find device with uuid ylGuQS-L7Xi-7nu7-ZHfA-ECba-8t4n-qIsvqc.
  Couldn't find device with uuid XOop3m-0Mw7-4Ysx-vjxD-xjbd-RKi2-XqL3ba.
  Couldn't find device with uuid GQDoib-Uus6-N1B9-Rk0B-tULK-XKc0-856QO9.
  Couldn't find device with uuid UW3TaW-tkKd-N5sy-Z8dD-S8wl-362C-ssTZ7j.
  Cannot restore Volume Group chenbro0.0 with 7 PVs marked as missing.
  Restore failed.

Would you know if possible and if so then how to recover from such a
failure?

many thanks, L


[-- Attachment #2: pEpkey.asc --]
[-- Type: application/pgp-keys, Size: 1787 bytes --]

Re: [linux-lvm] recovery of a VG - how?

[David Teigland]

On Mon, Oct 07, 2019 at 05:25:15PM +0100, lejeczek wrote:
> hi guys
> 
> I have a VG which I try to restore:
> 
> $ vgcfgrestore --file /etc/lvm/archive/chenbro0.0_00308-2121179640.vg chenbro0.0
> � /dev/mapper/chenbro_a: Checksum error at offset 428544
> � Couldn't read volume group metadata from /dev/mapper/chenbro_a.
> � Metadata location on /dev/mapper/chenbro_a at 428544 has invalid summary for VG.

The latest upstream version of lvm has some improvements to help recover
from problems like this.  So, you could try building lvm 2.03.05, and
running 'vgck --updatemetadata'.  Otherwise, the old approach has been to
use pvcreate with --restorefile and --uuid, followed by vgcfgrestore.
(I'd experiment with that using some test devices if possible.)
Dave

Re: [linux-lvm] recovery of a VG - how?

[lejeczek]

[-- Attachment #1: Type: text/plain, Size: 916 bytes --]

On 07/10/2019 17:42, David Teigland wrote:
> On Mon, Oct 07, 2019 at 05:25:15PM +0100, lejeczek wrote:
>> hi guys
>>
>> I have a VG which I try to restore:
>>
>> $ vgcfgrestore --file /etc/lvm/archive/chenbro0.0_00308-2121179640.vg chenbro0.0
>>   /dev/mapper/chenbro_a: Checksum error at offset 428544
>>   Couldn't read volume group metadata from /dev/mapper/chenbro_a.
>>   Metadata location on /dev/mapper/chenbro_a at 428544 has invalid summary for VG.
> The latest upstream version of lvm has some improvements to help recover
> from problems like this.  So, you could try building lvm 2.03.05, and
> running 'vgck --updatemetadata'.  Otherwise, the old approach has been to
> use pvcreate with --restorefile and --uuid, followed by vgcfgrestore.
> (I'd experiment with that using some test devices if possible.)
> Dave

is that new version in rhel/centos 8 by any chance?

thanks, L.


[-- Attachment #2: pEpkey.asc --]
[-- Type: application/pgp-keys, Size: 1787 bytes --]

Re: [linux-lvm] recovery of a VG - how?

[lejeczek]

[-- Attachment #1: Type: text/plain, Size: 1366 bytes --]

On 07/10/2019 17:42, David Teigland wrote:
> On Mon, Oct 07, 2019 at 05:25:15PM +0100, lejeczek wrote:
>> hi guys
>>
>> I have a VG which I try to restore:
>>
>> $ vgcfgrestore --file /etc/lvm/archive/chenbro0.0_00308-2121179640.vg chenbro0.0
>>   /dev/mapper/chenbro_a: Checksum error at offset 428544
>>   Couldn't read volume group metadata from /dev/mapper/chenbro_a.
>>   Metadata location on /dev/mapper/chenbro_a at 428544 has invalid summary for VG.
> The latest upstream version of lvm has some improvements to help recover
> from problems like this.  So, you could try building lvm 2.03.05, and
> running 'vgck --updatemetadata'.  Otherwise, the old approach has been to
> use pvcreate with --restorefile and --uuid, followed by vgcfgrestore.
> (I'd experiment with that using some test devices if possible.)
> Dave
>
> _______________________________________________
> 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/

also, is it possible to stop monitoring, or stop whatever is responsible
for flooding the logs with:

...

lrmd[14636]:  notice: chenbro0.1-raid5_monitor_20000:419255:stderr [  
Couldn't read volume group metadata from /dev/mapper/chenbro_m. ]

...

for a time being?

many thanks, L.


[-- Attachment #2: pEpkey.asc --]
[-- Type: application/pgp-keys, Size: 1787 bytes --]