vgscan does not found my VG, but pvscan does...
Hi *,
I have a strage problem on my Desktop. I have installed Kernel 2.2.18 and lvm 0.9.
All things have been fine till this morning. I have increased a logical volume (and the reiserfs which resides on that volume) from 1000M to 2000M (not 2GB).
Some time later I had to reboot the machine (because I want to attach some devices to the SCSI bus). After the reboot vgscan does NOT find my VG any longer. It only says something like "no voulme group found".
If I call pvscan it tells me that it found two physical disks (thats ok). But both PVs are in an unknown VG. I should call vgscan (but see obove :-)
Then I called pvdata to find out something out about whats going wrong here.
pvdata -a /dev/sda1 ; pvdata -a /dev/sda2 ; are working fine , they also tell me that these PVs are members of sysvg.
Does anybody have an idear how to get vgscan to read this info? pvdata finds members of a VG but vgscan does not :-((.
I hope someone can help me!!
(Yes I do have backup data, but I want to sleep well, if there are so many lvm installations out in the world).
Best regards,
Fabian Herschel
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Have you tried a vgimport?
- -----Original Message-----
From: [email protected] [mailto:[email protected]]On
Behalf Of fabian herschel
Sent: Tuesday, 24 April 2001 13:01
To: [email protected]
Cc: [email protected]
Subject: [linux-lvm] vgscan does not found my VG, but pvscan does...
Hi *,
I have a strage problem on my Desktop. I have installed Kernel 2.2.18 and lvm 0.9.
All things have been fine till this morning. I have increased a logical volume
(and the reiserfs which resides on that volume) from 1000M to 2000M (not 2GB).
Some time later I had to reboot the machine (because I want to attach some
devices to the SCSI bus). After the reboot vgscan does NOT find my VG any
longer. It only says something like "no voulme group found".
If I call pvscan it tells me that it found two physical disks (thats ok). But
both PVs are in an unknown VG. I should call vgscan (but see obove :-)
Then I called pvdata to find out something out about whats going wrong here.
pvdata -a /dev/sda1 ; pvdata -a /dev/sda2 ; are working fine , they also tell
me that these PVs are members of sysvg.
Does anybody have an idear how to get vgscan to read this info? pvdata finds members of a VG but vgscan does not :-((.
I hope someone can help me!!
(Yes I do have backup data, but I want to sleep well, if there are so many
lvm installations out in the world).
Best regards,
Fabian Herschel
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
-----BEGIN PGP SIGNATURE-----
Version: PGP 7.0.1
iQA/AwUBOuZCcXC+DNfF0nVpEQJxIQCgjl8vznkeomiBLbB6xO5GBN8EOqMAoJRH +cW+M//7QOL0tJhHc3//KDGr
=0uO4
-----END PGP SIGNATURE-----
Quoted 4 lines · you
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1Have you tried a vgimport?
Yes I did. vgimport does not want to import the VG, because it is not an exported one :((.
On an other machine I had the same problem and I tried to make a vgexport. The _export_ went well (the VG was exported but I could not re-import it ).
I tried some things last night...
- First I have saved the heder blocks (VGDA) with dd (dd if=/dev/sda1 \
of=/root/HEAD_sda1 count=4596 bs=512)
- Then I called a vgcfgrestore -n sysvg /dev/sda1.
- Now pvscan tells me both PVs are in the VG sysvg.
- "vgchange -ay sysvg" activated the VG and the LVs could me mounted
BUT
- vgscan does NOT find the sysvg!!!
- After a reboot the sysvg is not available.
- Every time I have to call a vgcfgrestore :((
I have written a C-program which could retrieve all data from a LV, if pvdata -E reports correctly, but the LV could not be activated. But this is only a very bad work arround to retrieve my data. But I need a way to reactivate my VG, so it could be found by vgscan.
I'm hopeful we will find a solution.
Fabian Herschel
Quoted 24 lines · fabian herschel
Quoted 4 lines · you
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1Have you tried a vgimport?
Yes I did. vgimport does not want to import the VG, because it is not an exported one :((.
On an other machine I had the same problem and I tried to make a vgexport. The _export_ went well (the VG was exported but I could not re-import it ).
I tried some things last night...
- First I have saved the heder blocks (VGDA) with dd (dd if=/dev/sda1 \of=/root/HEAD_sda1 count=4596 bs=512)- Then I called a vgcfgrestore -n sysvg
/dev/sda1.
- Now pvscan tells me both PVs are in the VG sysvg.
- "vgchange -ay sysvg" activated the VG and the LVs could me mountedBUT
- vgscan does NOT find the sysvg!!!
- After a reboot the sysvg is not available.
- Every time I have to call a vgcfgrestore :((
Quoted 6 lines
I have written a C-program which could retrieve all data from a LV, if pvdata -E reports correctly, but the LV could not be activated. But this is only a very bad work arround to retrieve my data. But I need a way to reactivate my VG, so it could be found by vgscan.
I'm hopeful we will find a solution.
These "vgscan can't find my..." errors still show up on a too regular base :-( They absolutely need to be fixed before the 1.0 release.
Could you please send your two VGDAs compressed email attached to mee in order to be able to investigate more.
Add the "vgscan -d" output for the fail case as well, please.
Quoted 5 lines
Fabian Herschel
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
Signature · 15 lines
Regards,
Heinz -- The LVM Guy --
*** Software bugs are stupid.
Nevertheless it needs not so stupid people to solve them ***
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Heinz Mauelshagen Sistina Software Inc.
Senior Consultant/Developer Am Sonnenhang 11
56242 Marienrachdorf
Germany
[email protected] +49 2626 141200
FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-Quoted 9 lines · Heinz J. Mauelshagen
Quoted 1 line
I'm hopeful we will find a solution.
These "vgscan can't find my..." errors still show up on a too regular base :-( They absolutely need to be fixed before the 1.0 release.
Could you please send your two VGDAs compressed email attached to mee in order to be able to investigate more.
Add the "vgscan -d" output for the fail case as well, please.
Are the first 5000 blocks of each PV enought?
Is the complete VGDA stored in this area?
Fabian
Signature · 4 lines
Fabian Herschel SuSE Linux Solutions AG Tel: +49-6196-50951-23 Mergenthalerallee 45-47 Fax: +49-6196-409-607 D-65760 Eschborn mailto: [email protected] http://www.suse.de
Quoted 23 lines · [email protected] wrote:
Quoted 9 lines · Heinz J. Mauelshagen
Quoted 1 line
I'm hopeful we will find a solution.
These "vgscan can't find my..." errors still show up on a too regular base :-( They absolutely need to be fixed before the 1.0 release.
Could you please send your two VGDAs compressed email attached to mee in order to be able to investigate more.
Add the "vgscan -d" output for the fail case as well, please.
Are the first 5000 blocks of each PV enought?
Is the complete VGDA stored in this area?Fabian
--
Fabian Herschel SuSE Linux Solutions AG Tel: +49-6196-50951-23 Mergenthalerallee 45-47 Fax: +49-6196-409-607 D-65760 Eschborn mailto: [email protected] http://www.suse.de
Signature · 4 lines
Fabian Herschel SuSE Linux Solutions AG Tel: +49-6196-50951-23 Mergenthalerallee 45-47 Fax: +49-6196-409-607 D-65760 Eschborn mailto: [email protected] http://www.suse.de
Quoted 7 lines · On Wed, Apr 25, 2001 at 02:18:51PM +0200, [email protected] wrote:
Quoted 5 lines · Heinz J. Mauelshagen
Quoted 1 line
I'm hopeful we will find a solution.
These "vgscan can't find my..." errors still show up on a too regular base :-( They absolutely need to be fixed before the 1.0 release.
Quoted 6 lines
Quoted 4 lines
Could you please send your two VGDAs compressed email attached to mee in order to be able to investigate more.
Add the "vgscan -d" output for the fail case as well, please.
Are the first 5000 blocks of each PV enought?
That's ok.
Quoted 1 line
Is the complete VGDA stored in this area?
Typically yes, because the VGDA is smaller than a PE which is 4MB in size by default.
Quoted 14 lines
Fabian
Signature · 8 lines
Fabian Herschel SuSE Linux Solutions AG Tel: +49-6196-50951-23 Mergenthalerallee 45-47 Fax: +49-6196-409-607 D-65760 Eschborn mailto: [email protected] http://www.suse.de _______________________________________________ linux-lvm mailing list [email protected] http://lists.sistina.com/mailman/listinfo/linux-lvm
Signature · 15 lines
Regards,
Heinz -- The LVM Guy --
*** Software bugs are stupid.
Nevertheless it needs not so stupid people to solve them ***
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Heinz Mauelshagen Sistina Software Inc.
Senior Consultant/Developer Am Sonnenhang 11
56242 Marienrachdorf
Germany
[email protected] +49 2626 141200
FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-Quoted 11 lines · Fabian Herschel
I tried some things last night...
- First I have saved the heder blocks (VGDA) with dd (dd if=/dev/sda1 \of=/root/HEAD_sda1 count=4596 bs=512)- Then I called a vgcfgrestore -n sysvg
/dev/sda1.
- Now pvscan tells me both PVs are in the VG sysvg.
- "vgchange -ay sysvg" activated the VG and the LVs could me mountedBUT
- vgscan does NOT find the sysvg!!!
- After a reboot the sysvg is not available.
- Every time I have to call a vgcfgrestore :((
Can you post "pvdata -PPU <dev>" for each PV in the VG? Also, can you look through "vgscan -v -d" output to see why it is not finding all of the PVs it wants? This sounds like the old UUID problem.
Cheers, Andreas
Signature · 3 lines
Andreas Dilger \ "If a man ate a pound of pasta and a pound of antipasto,
\ would they cancel out, leaving him still hungry?"
http://www-mddsp.enel.ucalgary.ca/People/adilger/ -- DogbertQuoted 9 lines · Andreas Dilger
Quoted 7 lines · Fabian Herschel
I tried some things last night...
- First I have saved the heder blocks (VGDA) with dd (dd if=/dev/sda1 \of=/root/HEAD_sda1 count=4596 bs=512)- Then I called a vgcfgrestore -n sysvg
/dev/sda1.
- Now pvscan tells me both PVs are in the VG sysvg.
- "vgchange -ay sysvg" activated the VG and the LVs could me mounted
Quoted 8 lines
Quoted 4 lines
BUT
- vgscan does NOT find the sysvg!!!
- After a reboot the sysvg is not available.
- Every time I have to call a vgcfgrestore :((Can you post "pvdata -PPU <dev>" for each PV in the VG? Also, can you look through "vgscan -v -d" output to see why it is not finding all of the PVs it wants? This sounds like the old UUID problem.
Yes I will post that. I have to collect this data at home :)). I also will have a look up, if the tips about the UUID problem fit to my problem /reading the several pvdata cmd outs.
Thanks,
Fabian
Quoted 10 lines
Cheers, Andreas
--
Andreas Dilger \ "If a man ate a pound of pasta and a pound of antipasto,
\ would they cancel out, leaving him still hungry?"
http://www-mddsp.enel.ucalgary.ca/People/adilger/ -- Dogbert
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
Signature · 4 lines
Fabian Herschel SuSE Linux Solutions AG Tel: +49-6196-50951-23 Mergenthalerallee 45-47 Fax: +49-6196-409-607 D-65760 Eschborn mailto: [email protected] http://www.suse.de