[linux-lvm] vg gone after vgextend

[linux-lvm] vg gone after vgextend

[Eli Criffield]

[-- Attachment #1: Type: text/plain, Size: 2323 bytes --]

Hello

I had LVM up and working great, even on the / filesystem. I had a
/dev/hda1 that was a 4 gig filesystem that i tared my distribution to and
/dev/hda2 that use to be my / filesystem and /dev/hda3 that was my volume
group. I resized /dev/hda1 down to around 30 megs and made /dev/hda2
4+gig. I made /dev/hda1 /boot and had a initrd to turn on LVM and boot to
the / filesystem on /dev/rootvg/root_lv. That all worked fine.

Until i added /dev/hda2 (now about 4gig) to the volumegroup rootvg. All
looked ok when i went to reboot. On the reboot it died in the ramdisk when
it ran vgscan.
"vgscan -- no volume groups found"

I was then stuck with a none bootable machine. I'll save you the long
story about the trouble i had making some kind of rescue disk, but i
finally got one to work off a cd i had a friend burn for me.

Now when i run pvscan i get 
"pvscan -- inactive PV "/dev/hda3" is associated to an unknown VG (run
vgscan)
pvscan -- total: 14.4GB / in use 1 [14.4GB] / in no VG 0 [0]"

it doesn't see /dev/hda2 that i added to the vg right before reboot, but
thats not a problem because there was nothing on it yet anyway.

pvscan -vv gives

-- Physical volume ---
PV Name		/dev/hda3
VG Name		rootvg
PV Size		14.4 GB / NOT usable 3.48 MB [LVM: 135KB]
PV#		1
PV Status	available
Allocatable	yes
CurLV		6
P ESize (KByte)	4096
Total PE	3685
Free PE		1261
Allocated PE	2424
PV UUID		elhLo-Mw5J-8qaE-gN2D-zEYo-Woqh-qH7H5a
System Id	eli.custom-edge.com993738446


from what i can tell from the archives to do a vgcfgrestore is the next
logical thing but it doesn't work for me

vgcfgrestore -n rootvg /dev/hda3
vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist
vgcfgrestore -- a backup for volume group rootvg is not available

which makes since because i booted off a rescue disk in a ram disk.

vgscan still gives
vgscan -- no volume groups found

i did find this answer in the archives for a similar problem:

> You need to edit the disk signatures and for each PV, insert all the
> _other_ PV signatures in the right place.

but no explanation on how to do that?

Is that what i should be doing? If so how do i do that? If not what can i
do to save this box?

thanks a ton it will save a lot of hassle if i don't have to reinstall.

eli

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[AJ Lewis]

[-- Attachment #1: Type: text/plain, Size: 3338 bytes --]

Goto ftp://ftp.sistina.com/pub/LVM/contrib and get
uuid_fixer-0.3-IOP10.tar.gz

Read the README.  Run it on your system, then run vgscan.  It should fix
your problem.  If not, let me know.

On Tue, Jul 10, 2001 at 09:39:02AM -0500, Eli Criffield wrote:
> I had LVM up and working great, even on the / filesystem. I had a
> /dev/hda1 that was a 4 gig filesystem that i tared my distribution to and
> /dev/hda2 that use to be my / filesystem and /dev/hda3 that was my volume
> group. I resized /dev/hda1 down to around 30 megs and made /dev/hda2
> 4+gig. I made /dev/hda1 /boot and had a initrd to turn on LVM and boot to
> the / filesystem on /dev/rootvg/root_lv. That all worked fine.
> 
> Until i added /dev/hda2 (now about 4gig) to the volumegroup rootvg. All
> looked ok when i went to reboot. On the reboot it died in the ramdisk when
> it ran vgscan.
> "vgscan -- no volume groups found"
> 
> I was then stuck with a none bootable machine. I'll save you the long
> story about the trouble i had making some kind of rescue disk, but i
> finally got one to work off a cd i had a friend burn for me.
> 
> Now when i run pvscan i get 
> "pvscan -- inactive PV "/dev/hda3" is associated to an unknown VG (run
> vgscan)
> pvscan -- total: 14.4GB / in use 1 [14.4GB] / in no VG 0 [0]"
> 
> it doesn't see /dev/hda2 that i added to the vg right before reboot, but
> thats not a problem because there was nothing on it yet anyway.
> 
> pvscan -vv gives
> 
> -- Physical volume ---
> PV Name		/dev/hda3
> VG Name		rootvg
> PV Size		14.4 GB / NOT usable 3.48 MB [LVM: 135KB]
> PV#		1
> PV Status	available
> Allocatable	yes
> CurLV		6
> P ESize (KByte)	4096
> Total PE	3685
> Free PE		1261
> Allocated PE	2424
> PV UUID		elhLo-Mw5J-8qaE-gN2D-zEYo-Woqh-qH7H5a
> System Id	eli.custom-edge.com993738446
> 
> 
> from what i can tell from the archives to do a vgcfgrestore is the next
> logical thing but it doesn't work for me
> 
> vgcfgrestore -n rootvg /dev/hda3
> vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist
> vgcfgrestore -- a backup for volume group rootvg is not available
> 
> which makes since because i booted off a rescue disk in a ram disk.
> 
> vgscan still gives
> vgscan -- no volume groups found
> 
> i did find this answer in the archives for a similar problem:
> 
> > You need to edit the disk signatures and for each PV, insert all the
> > _other_ PV signatures in the right place.
> 
> but no explanation on how to do that?
> 
> Is that what i should be doing? If so how do i do that? If not what can i
> do to save this box?
> 
> thanks a ton it will save a lot of hassle if i don't have to reinstall.
> 
> eli



-- 
AJ Lewis
Sistina Software Inc.                  Voice:  612-638-0500
1313 5th St SE, Suite 111              Fax:    612-638-0500
Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
http://www.sistina.com

Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
Get my key at: http://www.sistina.com/~lewis/gpgkey
 (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)

-----Begin Obligatory Humorous Quote----------------------------------------
Hey! It compiles! Ship it!
-----End Obligatory Humorous Quote------------------------------------------

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[Eli Criffield]

[-- Attachment #1: Type: text/plain, Size: 4086 bytes --]

It didn't work :(

when i ran 
./uuid_fixer /dev/hda2 /dev/hda3    i got
/dev/hda2 - pv_read{}: PV identifier invalid

but if i run 
./uuid_fixer /dev/hda3 i get
Error: number of PVs passed in does not match number of PVs in /dev/hda3s
VG
1 PVs where passed in and 2 where expected

So it thinks /dev/hda2 is part of the vg but /dev/hda2 had no info on it
saying that it is.

oh btw im using LVM 0.9.1_beta2 as it came with kernel 2.4.5 and LVM
tools 0.9.1_beta2.

Now what?

eli

On Tue, Jul 10, 2001 at 11:06:56AM -0500, AJ Lewis wrote:
> Goto ftp://ftp.sistina.com/pub/LVM/contrib and get
> uuid_fixer-0.3-IOP10.tar.gz
> 
> Read the README.  Run it on your system, then run vgscan.  It should fix
> your problem.  If not, let me know.
> 
> On Tue, Jul 10, 2001 at 09:39:02AM -0500, Eli Criffield wrote:
> > I had LVM up and working great, even on the / filesystem. I had a
> > /dev/hda1 that was a 4 gig filesystem that i tared my distribution to and
> > /dev/hda2 that use to be my / filesystem and /dev/hda3 that was my volume
> > group. I resized /dev/hda1 down to around 30 megs and made /dev/hda2
> > 4+gig. I made /dev/hda1 /boot and had a initrd to turn on LVM and boot to
> > the / filesystem on /dev/rootvg/root_lv. That all worked fine.
> > 
> > Until i added /dev/hda2 (now about 4gig) to the volumegroup rootvg. All
> > looked ok when i went to reboot. On the reboot it died in the ramdisk when
> > it ran vgscan.
> > "vgscan -- no volume groups found"
> > 
> > I was then stuck with a none bootable machine. I'll save you the long
> > story about the trouble i had making some kind of rescue disk, but i
> > finally got one to work off a cd i had a friend burn for me.
> > 
> > Now when i run pvscan i get 
> > "pvscan -- inactive PV "/dev/hda3" is associated to an unknown VG (run
> > vgscan)
> > pvscan -- total: 14.4GB / in use 1 [14.4GB] / in no VG 0 [0]"
> > 
> > it doesn't see /dev/hda2 that i added to the vg right before reboot, but
> > thats not a problem because there was nothing on it yet anyway.
> > 
> > pvscan -vv gives
> > 
> > -- Physical volume ---
> > PV Name		/dev/hda3
> > VG Name		rootvg
> > PV Size		14.4 GB / NOT usable 3.48 MB [LVM: 135KB]
> > PV#		1
> > PV Status	available
> > Allocatable	yes
> > CurLV		6
> > P ESize (KByte)	4096
> > Total PE	3685
> > Free PE		1261
> > Allocated PE	2424
> > PV UUID		elhLo-Mw5J-8qaE-gN2D-zEYo-Woqh-qH7H5a
> > System Id	eli.custom-edge.com993738446
> > 
> > 
> > from what i can tell from the archives to do a vgcfgrestore is the next
> > logical thing but it doesn't work for me
> > 
> > vgcfgrestore -n rootvg /dev/hda3
> > vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist
> > vgcfgrestore -- a backup for volume group rootvg is not available
> > 
> > which makes since because i booted off a rescue disk in a ram disk.
> > 
> > vgscan still gives
> > vgscan -- no volume groups found
> > 
> > i did find this answer in the archives for a similar problem:
> > 
> > > You need to edit the disk signatures and for each PV, insert all the
> > > _other_ PV signatures in the right place.
> > 
> > but no explanation on how to do that?
> > 
> > Is that what i should be doing? If so how do i do that? If not what can i
> > do to save this box?
> > 
> > thanks a ton it will save a lot of hassle if i don't have to reinstall.
> > 
> > eli
> 
> 
> 
> -- 
> AJ Lewis
> Sistina Software Inc.                  Voice:  612-638-0500
> 1313 5th St SE, Suite 111              Fax:    612-638-0500
> Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
> http://www.sistina.com
> 
> Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
> Get my key at: http://www.sistina.com/~lewis/gpgkey
>  (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)
> 
> -----Begin Obligatory Humorous Quote----------------------------------------
> Hey! It compiles! Ship it!
> -----End Obligatory Humorous Quote------------------------------------------



[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[AJ Lewis]

[-- Attachment #1: Type: text/plain, Size: 5282 bytes --]

On Tue, Jul 10, 2001 at 11:39:17AM -0500, Eli Criffield wrote:
> It didn't work :(
> 
> when i ran 
> ./uuid_fixer /dev/hda2 /dev/hda3    i got
> /dev/hda2 - pv_read{}: PV identifier invalid

are you *sure* that /dev/hda2 is the other PV?  What happens when you run
'pvdata -U /dev/hda2'?  You might also try recompiling the uuid_fixer
against the beta2 source, and running it again.  It should work as is, but
that might be an issue.

> but if i run 
> ./uuid_fixer /dev/hda3 i get
> Error: number of PVs passed in does not match number of PVs in /dev/hda3s
> VG
> 1 PVs where passed in and 2 where expected
> 
> So it thinks /dev/hda2 is part of the vg but /dev/hda2 had no info on it
> saying that it is.
> 
> oh btw im using LVM 0.9.1_beta2 as it came with kernel 2.4.5 and LVM
> tools 0.9.1_beta2.
> 
> Now what?
> 
> eli
> 
> On Tue, Jul 10, 2001 at 11:06:56AM -0500, AJ Lewis wrote:
> > Goto ftp://ftp.sistina.com/pub/LVM/contrib and get
> > uuid_fixer-0.3-IOP10.tar.gz
> > 
> > Read the README.  Run it on your system, then run vgscan.  It should fix
> > your problem.  If not, let me know.
> > 
> > On Tue, Jul 10, 2001 at 09:39:02AM -0500, Eli Criffield wrote:
> > > I had LVM up and working great, even on the / filesystem. I had a
> > > /dev/hda1 that was a 4 gig filesystem that i tared my distribution to and
> > > /dev/hda2 that use to be my / filesystem and /dev/hda3 that was my volume
> > > group. I resized /dev/hda1 down to around 30 megs and made /dev/hda2
> > > 4+gig. I made /dev/hda1 /boot and had a initrd to turn on LVM and boot to
> > > the / filesystem on /dev/rootvg/root_lv. That all worked fine.
> > > 
> > > Until i added /dev/hda2 (now about 4gig) to the volumegroup rootvg. All
> > > looked ok when i went to reboot. On the reboot it died in the ramdisk when
> > > it ran vgscan.
> > > "vgscan -- no volume groups found"
> > > 
> > > I was then stuck with a none bootable machine. I'll save you the long
> > > story about the trouble i had making some kind of rescue disk, but i
> > > finally got one to work off a cd i had a friend burn for me.
> > > 
> > > Now when i run pvscan i get 
> > > "pvscan -- inactive PV "/dev/hda3" is associated to an unknown VG (run
> > > vgscan)
> > > pvscan -- total: 14.4GB / in use 1 [14.4GB] / in no VG 0 [0]"
> > > 
> > > it doesn't see /dev/hda2 that i added to the vg right before reboot, but
> > > thats not a problem because there was nothing on it yet anyway.
> > > 
> > > pvscan -vv gives
> > > 
> > > -- Physical volume ---
> > > PV Name		/dev/hda3
> > > VG Name		rootvg
> > > PV Size		14.4 GB / NOT usable 3.48 MB [LVM: 135KB]
> > > PV#		1
> > > PV Status	available
> > > Allocatable	yes
> > > CurLV		6
> > > P ESize (KByte)	4096
> > > Total PE	3685
> > > Free PE		1261
> > > Allocated PE	2424
> > > PV UUID		elhLo-Mw5J-8qaE-gN2D-zEYo-Woqh-qH7H5a
> > > System Id	eli.custom-edge.com993738446
> > > 
> > > 
> > > from what i can tell from the archives to do a vgcfgrestore is the next
> > > logical thing but it doesn't work for me
> > > 
> > > vgcfgrestore -n rootvg /dev/hda3
> > > vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist
> > > vgcfgrestore -- a backup for volume group rootvg is not available
> > > 
> > > which makes since because i booted off a rescue disk in a ram disk.
> > > 
> > > vgscan still gives
> > > vgscan -- no volume groups found
> > > 
> > > i did find this answer in the archives for a similar problem:
> > > 
> > > > You need to edit the disk signatures and for each PV, insert all the
> > > > _other_ PV signatures in the right place.
> > > 
> > > but no explanation on how to do that?
> > > 
> > > Is that what i should be doing? If so how do i do that? If not what can i
> > > do to save this box?
> > > 
> > > thanks a ton it will save a lot of hassle if i don't have to reinstall.
> > > 
> > > eli
> > 
> > 
> > 
> > -- 
> > AJ Lewis
> > Sistina Software Inc.                  Voice:  612-638-0500
> > 1313 5th St SE, Suite 111              Fax:    612-638-0500
> > Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
> > http://www.sistina.com
> > 
> > Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
> > Get my key at: http://www.sistina.com/~lewis/gpgkey
> >  (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)
> > 
> > -----Begin Obligatory Humorous Quote----------------------------------------
> > Hey! It compiles! Ship it!
> > -----End Obligatory Humorous Quote------------------------------------------
> 
> 



-- 
AJ Lewis
Sistina Software Inc.                  Voice:  612-638-0500
1313 5th St SE, Suite 111              Fax:    612-638-0500
Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
http://www.sistina.com

Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
Get my key at: http://www.sistina.com/~lewis/gpgkey
 (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)

-----Begin Obligatory Humorous Quote----------------------------------------
Computers make very fast, very accurate mistakes.
-----End Obligatory Humorous Quote------------------------------------------

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[Eli Criffield]

[-- Attachment #1: Type: text/plain, Size: 6163 bytes --]

Yeah /dev/hda2 is the pv i added to rootvg just before the reboot.
I only have 3 partitions hda1 is root, hda2 LVM and hda3 is LVM.

I recompiled uuid_fixer and tried it again, but got the exact same
results.

pvdata -U /dev/hda2 returns
pvdata -- ERROR "pv_read{}: PV identifier invalid" reading phsical volume
uuid list from physical volume "/dev/hda2"

pvdata -U /dev/hda3 gives
--- List of phsical volume UUIDs ---

000: elhLXoMw5J8qaEgN2DzEYoWoqhqH7H5a
001: 9AoOf5F6c58dcyJ30pfwNp153TI5PjLF


thanks again

eli


On Tue, Jul 10, 2001 at 12:12:06PM -0500, AJ Lewis wrote:
> On Tue, Jul 10, 2001 at 11:39:17AM -0500, Eli Criffield wrote:
> > It didn't work :(
> > 
> > when i ran 
> > ./uuid_fixer /dev/hda2 /dev/hda3    i got
> > /dev/hda2 - pv_read{}: PV identifier invalid
> 
> are you *sure* that /dev/hda2 is the other PV?  What happens when you run
> 'pvdata -U /dev/hda2'?  You might also try recompiling the uuid_fixer
> against the beta2 source, and running it again.  It should work as is, but
> that might be an issue.
> 
> > but if i run 
> > ./uuid_fixer /dev/hda3 i get
> > Error: number of PVs passed in does not match number of PVs in /dev/hda3s
> > VG
> > 1 PVs where passed in and 2 where expected
> > 
> > So it thinks /dev/hda2 is part of the vg but /dev/hda2 had no info on it
> > saying that it is.
> > 
> > oh btw im using LVM 0.9.1_beta2 as it came with kernel 2.4.5 and LVM
> > tools 0.9.1_beta2.
> > 
> > Now what?
> > 
> > eli
> > 
> > On Tue, Jul 10, 2001 at 11:06:56AM -0500, AJ Lewis wrote:
> > > Goto ftp://ftp.sistina.com/pub/LVM/contrib and get
> > > uuid_fixer-0.3-IOP10.tar.gz
> > > 
> > > Read the README.  Run it on your system, then run vgscan.  It should fix
> > > your problem.  If not, let me know.
> > > 
> > > On Tue, Jul 10, 2001 at 09:39:02AM -0500, Eli Criffield wrote:
> > > > I had LVM up and working great, even on the / filesystem. I had a
> > > > /dev/hda1 that was a 4 gig filesystem that i tared my distribution to and
> > > > /dev/hda2 that use to be my / filesystem and /dev/hda3 that was my volume
> > > > group. I resized /dev/hda1 down to around 30 megs and made /dev/hda2
> > > > 4+gig. I made /dev/hda1 /boot and had a initrd to turn on LVM and boot to
> > > > the / filesystem on /dev/rootvg/root_lv. That all worked fine.
> > > > 
> > > > Until i added /dev/hda2 (now about 4gig) to the volumegroup rootvg. All
> > > > looked ok when i went to reboot. On the reboot it died in the ramdisk when
> > > > it ran vgscan.
> > > > "vgscan -- no volume groups found"
> > > > 
> > > > I was then stuck with a none bootable machine. I'll save you the long
> > > > story about the trouble i had making some kind of rescue disk, but i
> > > > finally got one to work off a cd i had a friend burn for me.
> > > > 
> > > > Now when i run pvscan i get 
> > > > "pvscan -- inactive PV "/dev/hda3" is associated to an unknown VG (run
> > > > vgscan)
> > > > pvscan -- total: 14.4GB / in use 1 [14.4GB] / in no VG 0 [0]"
> > > > 
> > > > it doesn't see /dev/hda2 that i added to the vg right before reboot, but
> > > > thats not a problem because there was nothing on it yet anyway.
> > > > 
> > > > pvscan -vv gives
> > > > 
> > > > -- Physical volume ---
> > > > PV Name		/dev/hda3
> > > > VG Name		rootvg
> > > > PV Size		14.4 GB / NOT usable 3.48 MB [LVM: 135KB]
> > > > PV#		1
> > > > PV Status	available
> > > > Allocatable	yes
> > > > CurLV		6
> > > > P ESize (KByte)	4096
> > > > Total PE	3685
> > > > Free PE		1261
> > > > Allocated PE	2424
> > > > PV UUID		elhLo-Mw5J-8qaE-gN2D-zEYo-Woqh-qH7H5a
> > > > System Id	eli.custom-edge.com993738446
> > > > 
> > > > 
> > > > from what i can tell from the archives to do a vgcfgrestore is the next
> > > > logical thing but it doesn't work for me
> > > > 
> > > > vgcfgrestore -n rootvg /dev/hda3
> > > > vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist
> > > > vgcfgrestore -- a backup for volume group rootvg is not available
> > > > 
> > > > which makes since because i booted off a rescue disk in a ram disk.
> > > > 
> > > > vgscan still gives
> > > > vgscan -- no volume groups found
> > > > 
> > > > i did find this answer in the archives for a similar problem:
> > > > 
> > > > > You need to edit the disk signatures and for each PV, insert all the
> > > > > _other_ PV signatures in the right place.
> > > > 
> > > > but no explanation on how to do that?
> > > > 
> > > > Is that what i should be doing? If so how do i do that? If not what can i
> > > > do to save this box?
> > > > 
> > > > thanks a ton it will save a lot of hassle if i don't have to reinstall.
> > > > 
> > > > eli
> > > 
> > > 
> > > 
> > > -- 
> > > AJ Lewis
> > > Sistina Software Inc.                  Voice:  612-638-0500
> > > 1313 5th St SE, Suite 111              Fax:    612-638-0500
> > > Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
> > > http://www.sistina.com
> > > 
> > > Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
> > > Get my key at: http://www.sistina.com/~lewis/gpgkey
> > >  (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)
> > > 
> > > -----Begin Obligatory Humorous Quote----------------------------------------
> > > Hey! It compiles! Ship it!
> > > -----End Obligatory Humorous Quote------------------------------------------
> > 
> > 
> 
> 
> 
> -- 
> AJ Lewis
> Sistina Software Inc.                  Voice:  612-638-0500
> 1313 5th St SE, Suite 111              Fax:    612-638-0500
> Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
> http://www.sistina.com
> 
> Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
> Get my key at: http://www.sistina.com/~lewis/gpgkey
>  (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)
> 
> -----Begin Obligatory Humorous Quote----------------------------------------
> Computers make very fast, very accurate mistakes.
> -----End Obligatory Humorous Quote------------------------------------------



[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[AJ Lewis]

[-- Attachment #1: Type: text/plain, Size: 7448 bytes --]

On Tue, Jul 10, 2001 at 01:57:32PM -0500, Eli Criffield wrote:
> Yeah /dev/hda2 is the pv i added to rootvg just before the reboot.
> I only have 3 partitions hda1 is root, hda2 LVM and hda3 is LVM.
> 
> I recompiled uuid_fixer and tried it again, but got the exact same
> results.
> 
> pvdata -U /dev/hda2 returns
> pvdata -- ERROR "pv_read{}: PV identifier invalid" reading phsical volume
> uuid list from physical volume "/dev/hda2"

Hmm...have you tried 'vgcfgrestore'?  It looks like you managed to corrupt
the PV somehow...without the exact steps you took to get to this state, it's
gonna be hard to determine what happened.  LVM 0.9.1 beta2 has some pretty
odd problems.

> pvdata -U /dev/hda3 gives
> --- List of phsical volume UUIDs ---
> 
> 000: elhLXoMw5J8qaEgN2DzEYoWoqhqH7H5a
> 001: 9AoOf5F6c58dcyJ30pfwNp153TI5PjLF
> 
> 
> thanks again
> 
> eli
> 
> 
> On Tue, Jul 10, 2001 at 12:12:06PM -0500, AJ Lewis wrote:
> > On Tue, Jul 10, 2001 at 11:39:17AM -0500, Eli Criffield wrote:
> > > It didn't work :(
> > > 
> > > when i ran 
> > > ./uuid_fixer /dev/hda2 /dev/hda3    i got
> > > /dev/hda2 - pv_read{}: PV identifier invalid
> > 
> > are you *sure* that /dev/hda2 is the other PV?  What happens when you run
> > 'pvdata -U /dev/hda2'?  You might also try recompiling the uuid_fixer
> > against the beta2 source, and running it again.  It should work as is, but
> > that might be an issue.
> > 
> > > but if i run 
> > > ./uuid_fixer /dev/hda3 i get
> > > Error: number of PVs passed in does not match number of PVs in /dev/hda3s
> > > VG
> > > 1 PVs where passed in and 2 where expected
> > > 
> > > So it thinks /dev/hda2 is part of the vg but /dev/hda2 had no info on it
> > > saying that it is.
> > > 
> > > oh btw im using LVM 0.9.1_beta2 as it came with kernel 2.4.5 and LVM
> > > tools 0.9.1_beta2.
> > > 
> > > Now what?
> > > 
> > > eli
> > > 
> > > On Tue, Jul 10, 2001 at 11:06:56AM -0500, AJ Lewis wrote:
> > > > Goto ftp://ftp.sistina.com/pub/LVM/contrib and get
> > > > uuid_fixer-0.3-IOP10.tar.gz
> > > > 
> > > > Read the README.  Run it on your system, then run vgscan.  It should fix
> > > > your problem.  If not, let me know.
> > > > 
> > > > On Tue, Jul 10, 2001 at 09:39:02AM -0500, Eli Criffield wrote:
> > > > > I had LVM up and working great, even on the / filesystem. I had a
> > > > > /dev/hda1 that was a 4 gig filesystem that i tared my distribution to and
> > > > > /dev/hda2 that use to be my / filesystem and /dev/hda3 that was my volume
> > > > > group. I resized /dev/hda1 down to around 30 megs and made /dev/hda2
> > > > > 4+gig. I made /dev/hda1 /boot and had a initrd to turn on LVM and boot to
> > > > > the / filesystem on /dev/rootvg/root_lv. That all worked fine.
> > > > > 
> > > > > Until i added /dev/hda2 (now about 4gig) to the volumegroup rootvg. All
> > > > > looked ok when i went to reboot. On the reboot it died in the ramdisk when
> > > > > it ran vgscan.
> > > > > "vgscan -- no volume groups found"
> > > > > 
> > > > > I was then stuck with a none bootable machine. I'll save you the long
> > > > > story about the trouble i had making some kind of rescue disk, but i
> > > > > finally got one to work off a cd i had a friend burn for me.
> > > > > 
> > > > > Now when i run pvscan i get 
> > > > > "pvscan -- inactive PV "/dev/hda3" is associated to an unknown VG (run
> > > > > vgscan)
> > > > > pvscan -- total: 14.4GB / in use 1 [14.4GB] / in no VG 0 [0]"
> > > > > 
> > > > > it doesn't see /dev/hda2 that i added to the vg right before reboot, but
> > > > > thats not a problem because there was nothing on it yet anyway.
> > > > > 
> > > > > pvscan -vv gives
> > > > > 
> > > > > -- Physical volume ---
> > > > > PV Name		/dev/hda3
> > > > > VG Name		rootvg
> > > > > PV Size		14.4 GB / NOT usable 3.48 MB [LVM: 135KB]
> > > > > PV#		1
> > > > > PV Status	available
> > > > > Allocatable	yes
> > > > > CurLV		6
> > > > > P ESize (KByte)	4096
> > > > > Total PE	3685
> > > > > Free PE		1261
> > > > > Allocated PE	2424
> > > > > PV UUID		elhLo-Mw5J-8qaE-gN2D-zEYo-Woqh-qH7H5a
> > > > > System Id	eli.custom-edge.com993738446
> > > > > 
> > > > > 
> > > > > from what i can tell from the archives to do a vgcfgrestore is the next
> > > > > logical thing but it doesn't work for me
> > > > > 
> > > > > vgcfgrestore -n rootvg /dev/hda3
> > > > > vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist
> > > > > vgcfgrestore -- a backup for volume group rootvg is not available
> > > > > 
> > > > > which makes since because i booted off a rescue disk in a ram disk.
> > > > > 
> > > > > vgscan still gives
> > > > > vgscan -- no volume groups found
> > > > > 
> > > > > i did find this answer in the archives for a similar problem:
> > > > > 
> > > > > > You need to edit the disk signatures and for each PV, insert all the
> > > > > > _other_ PV signatures in the right place.
> > > > > 
> > > > > but no explanation on how to do that?
> > > > > 
> > > > > Is that what i should be doing? If so how do i do that? If not what can i
> > > > > do to save this box?
> > > > > 
> > > > > thanks a ton it will save a lot of hassle if i don't have to reinstall.
> > > > > 
> > > > > eli
> > > > 
> > > > 
> > > > 
> > > > -- 
> > > > AJ Lewis
> > > > Sistina Software Inc.                  Voice:  612-638-0500
> > > > 1313 5th St SE, Suite 111              Fax:    612-638-0500
> > > > Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
> > > > http://www.sistina.com
> > > > 
> > > > Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
> > > > Get my key at: http://www.sistina.com/~lewis/gpgkey
> > > >  (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)
> > > > 
> > > > -----Begin Obligatory Humorous Quote----------------------------------------
> > > > Hey! It compiles! Ship it!
> > > > -----End Obligatory Humorous Quote------------------------------------------
> > > 
> > > 
> > 
> > 
> > 
> > -- 
> > AJ Lewis
> > Sistina Software Inc.                  Voice:  612-638-0500
> > 1313 5th St SE, Suite 111              Fax:    612-638-0500
> > Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
> > http://www.sistina.com
> > 
> > Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
> > Get my key at: http://www.sistina.com/~lewis/gpgkey
> >  (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)
> > 
> > -----Begin Obligatory Humorous Quote----------------------------------------
> > Computers make very fast, very accurate mistakes.
> > -----End Obligatory Humorous Quote------------------------------------------
> 
> 



-- 
AJ Lewis
Sistina Software Inc.                  Voice:  612-638-0500
1313 5th St SE, Suite 111              Fax:    612-638-0500
Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
http://www.sistina.com

Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
Get my key at: http://www.sistina.com/~lewis/gpgkey
 (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)

-----Begin Obligatory Humorous Quote----------------------------------------
APATHY ERROR: Don't bother striking any key.
-----End Obligatory Humorous Quote------------------------------------------

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[Eli Criffield]

[-- Attachment #1: Type: text/plain, Size: 526 bytes --]

> Hmm...have you tried 'vgcfgrestore'?  It looks like you managed to corrupt
> the PV somehow...without the exact steps you took to get to this state, it's
> gonna be hard to determine what happened.  LVM 0.9.1 beta2 has some pretty
> odd problems.

#vgcfgrestore -n rootvg /dev/hda3                                      
vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist              
vgcfgrestore -- a backup for volume group rootvg is not available     

Yeah tried that already? Is there any hope?

eli

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[AJ Lewis]

[-- Attachment #1: Type: text/plain, Size: 1273 bytes --]

On Tue, Jul 10, 2001 at 02:13:32PM -0500, Eli Criffield wrote:
> #vgcfgrestore -n rootvg /dev/hda3                                      
> vgcfgrestore -- "/etc/lvmconf/rootvg.conf" doesn't exist              
> vgcfgrestore -- a backup for volume group rootvg is not available     
> 
> Yeah tried that already? Is there any hope?

Have you looked in /etc/lvmconf?  Are there any conf files in there?  You
can specify the file to vgcfgrestore on the command line. If there isn't a
file there, don't give up hope yet.  Some of the folks on this list might
still respond with other options.

Regards
-- 
AJ Lewis
Sistina Software Inc.                  Voice:  612-638-0500
1313 5th St SE, Suite 111              Fax:    612-638-0500
Minneapolis, MN 55414                  E-Mail: lewis@sistina.com
http://www.sistina.com

Current GPG fingerprint = 3B5F 6011 5216 76A5 2F6B  52A0 941E 1261 0029 2648
Get my key at: http://www.sistina.com/~lewis/gpgkey
 (Unfortunately, the PKS-type keyservers do not work with multiple sub-keys)

-----Begin Obligatory Humorous Quote----------------------------------------
A program is a device used to convert data into error messages.
-----End Obligatory Humorous Quote------------------------------------------

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [linux-lvm] vg gone after vgextend

[Eli Criffield]

[-- Attachment #1: Type: text/plain, Size: 381 bytes --]

> Have you looked in /etc/lvmconf?  Are there any conf files in there?  You
> can specify the file to vgcfgrestore on the command line. If there isn't a
> file there, don't give up hope yet.  Some of the folks on this list might
> still respond with other options.

nope nothing in /etc/lvmconf, I'm booted to a rescuse disk and / is in a
ram disk.

Anybody Anybody Anybody??

eli

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]