Recovery of LVM after kernelupdate (2.4.10 -> 2.4.19)
Hello,
does anyone of you know, whether it is possible to restore an LVM just from the physical volumes and the VG.conf file ?
I did an update from Kernel 2.4.10 (SuSE 7.3) to 2.4.19 (SuSE 8.1) with the result that no volume groups are found :-(
My LVM structure is like this:
The only volume group is "hxvg01" which contains two volumes "hxlv01" (pv: hdc1, hdd1) and "hxlv02" (pv: hdb1)
I tried to use vgcfgrestore (@SuSE 8.1), but it failed with this messages:
vgcfgrestore -- INFO: using backup file "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- ERROR: different structure size stored in "/etc/lvmconf/hxvg01.conf" than expected in file vg_cfgrestore.c [line 139] vgcfgrestore -- ERROR "vg_cfgrestore(): read" restoring volume group "hxvg01"
Is this the often discussed "compatibility problem" ? How can I safe migrate to
the new LVM Version without loosing all my data OR (if it is not possible) how
can I access my LVM on a Linux 2.4.10 System ?
The vgcfgrestore-step taken on a minimal Linux (SuSE 7.3) with Kernel 2.4.10 shows that
following result. Which steps are necessary to access my volumes under Kernel 2.4.10 ?
Is it possible to import the VG using the -f parameter of vgimport ?
# vgcfgrestore -ll -n hxvg01 -t -v vgcfgrestore -- locking logical volume manager vgcfgrestore -- restoring volume group "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- checking existence of "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- reading volume group data for "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- reading physical volume data for "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- reading logical volume data for "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- checking volume group consistency of "hxvg01" vgcfgrestore -- checking volume group consistency of "hxvg01" vgcfgrestore -- backup of volume group "hxvg01" is consistent --- Volume group --- VG Name hxvg01 VG Access read/write VG Status NOT available/resizable VG # 0 MAX LV 256 Cur LV 2 Open LV 0 MAX LV Size 255.99 GB Max PV 256 Cur PV 3 Act PV 3 VG Size 160.53 GB PE Size 4.00 MB Total PE 41096 Alloc PE / Size 41075 / 160.45 GB Free PE / Size 21 / 84.00 MB VG UUID cXz5yh-KTaG-XRxy-v30t-ajxu-dV2W-Bpu1W0
--- Logical volume ---
LV Name /dev/hxvg01/hxlv01 VG Name hxvg01 LV Write Access read/write LV Status NOT available LV # 1 # open 0 LV Size 83.85 GB Current LE 21465 Allocated LE 21465 Allocation next free Read ahead sectors 128 Block device 58:0
--- Logical volume ---
LV Name /dev/hxvg01/hxlv02 VG Name hxvg01 LV Write Access read/write LV Status NOT available LV # 2 # open 0 LV Size 76.60 GB Current LE 19610 Allocated LE 19610 Allocation next free Read ahead sectors 120 Block device 58:1
--- Physical volume ---
PV Name /dev/hdd1 VG Name hxvg01 PV Size 41.93 GB / NOT usable 3.49 MB [LVM: 271.00 KB] PV# 1 PV Status available Allocatable yes (but full) Cur LV 1 PE Size (KByte) 4096 Total PE 10732 Free PE 0 Allocated PE 10732 PV UUID 5SaSBe-zk7g-3XR6-5X7a-H21x-3mW4-EJyHHN
--- Physical volume ---
PV Name /dev/hdc1 VG Name hxvg01 PV Size 41.93 GB / NOT usable 2.96 MB [LVM: 271.00 KB] PV# 2 PV Status available Allocatable yes (but full) Cur LV 1 PE Size (KByte) 4096 Total PE 10733 Free PE 0 Allocated PE 10733 PV UUID gQYeny-bpfF-IoQl-8dtG-867E-OXIj-FBqmlj
--- Physical volume ---
PV Name /dev/hdb1 VG Name hxvg01 PV Size 76.69 GB / NOT usable 4.64 MB [LVM: 308.00 KB] PV# 3 PV Status NOT available Allocatable yes Cur LV 1 PE Size (KByte) 4096 Total PE 19631 Free PE 21 Allocated PE 19610 PV UUID 3cQksh-0OAA-ar9H-3fms-AgGQ-tOIO-eREM8v
vgcfgrestore -- unlocking logical volume manager
Which of the pv* vg* lv* commands are changing the physical volume (=which are dangerous for me) ?
// melle
Signature · 2 lines
+++ GMX - Mail, Messaging & more http://www.gmx.net +++ NEU: Mit GMX ins Internet. Rund um die Uhr f�r 1 ct/ Min. surfen!
Quoted 7 lines · Thomas Mellenthin
Hello,
does anyone of you know, whether it is possible to restore an LVM just from the physical volumes and the VG.conf file ?
I did an update from Kernel 2.4.10 (SuSE 7.3) to 2.4.19 (SuSE 8.1) with the result that no volume groups are found :-(
Under SuSE 8.1:
- what does vgscan display ?
- what does pvscan display ? Does it show your PVs on hd[cd]1 ?
Quoted 6 lines
My LVM structure is like this:
The only volume group is "hxvg01" which contains two volumes "hxlv01" (pv: hdc1, hdd1) and "hxlv02" (pv: hdb1)I tried to use vgcfgrestore (@SuSE 8.1), but it failed with this messages:
No need to vgcfgrestore, because your metadata should still be on hd[cd]1, unless you (or yast) write somethig onto hd[cd]1. You didn't create filesystems during the update ther, did you ? ;) If you did -> full stop. No chance to retrieve anything particular usefull from the PVs again.
Key question is as asked above:
does pvscan (SuSE 8.1) find your 2 PVs on hd[cd]1 ?
Regards,
Heinz -- The LVM Guy --
Quoted 125 lines
vgcfgrestore -- INFO: using backup file "
/etc/lvmconf/hxvg01.conf" vgcfgrestore -- ERROR: different structure size stored in "/etc/lvmconf/hxvg01.conf" than expected in filevg_cfgrestore.c[line 139] vgcfgrestore -- ERROR "vg_cfgrestore(): read" restoring volume group "hxvg01"Is this the often discussed "compatibility problem" ? How can I safe migrate to
the new LVM Version without loosing all my data OR (if it is not possible) how
can I access my LVM on a Linux 2.4.10 System ?The vgcfgrestore-step taken on a minimal Linux (SuSE 7.3) with Kernel 2.4.10 shows that
following result. Which steps are necessary to access my volumes under Kernel 2.4.10 ?
Is it possible to import the VG using the -f parameter of vgimport ?# vgcfgrestore -ll -n hxvg01 -t -v vgcfgrestore -- locking logical volume manager vgcfgrestore -- restoring volume group "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- checking existence of "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- reading volume group data for "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- reading physical volume data for "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- reading logical volume data for "hxvg01" from "/etc/lvmconf/hxvg01.conf" vgcfgrestore -- checking volume group consistency of "hxvg01" vgcfgrestore -- checking volume group consistency of "hxvg01" vgcfgrestore -- backup of volume group "hxvg01" is consistent --- Volume group --- VG Name hxvg01 VG Access read/write VG Status NOT available/resizable VG # 0 MAX LV 256 Cur LV 2 Open LV 0 MAX LV Size 255.99 GB Max PV 256 Cur PV 3 Act PV 3 VG Size 160.53 GB PE Size 4.00 MB Total PE 41096 Alloc PE / Size 41075 / 160.45 GB Free PE / Size 21 / 84.00 MB VG UUID cXz5yh-KTaG-XRxy-v30t-ajxu-dV2W-Bpu1W0--- Logical volume ---
LV Name /dev/hxvg01/hxlv01 VG Name hxvg01 LV Write Access read/write LV Status NOT available LV # 1 # open 0 LV Size 83.85 GB Current LE 21465 Allocated LE 21465 Allocation next free Read ahead sectors 128 Block device 58:0--- Logical volume ---
LV Name /dev/hxvg01/hxlv02 VG Name hxvg01 LV Write Access read/write LV Status NOT available LV # 2 # open 0 LV Size 76.60 GB Current LE 19610 Allocated LE 19610 Allocation next free Read ahead sectors 120 Block device 58:1--- Physical volume ---
PV Name /dev/hdd1 VG Name hxvg01 PV Size 41.93 GB / NOT usable 3.49 MB [LVM: 271.00 KB] PV# 1 PV Status available Allocatable yes (but full) Cur LV 1 PE Size (KByte) 4096 Total PE 10732 Free PE 0 Allocated PE 10732 PV UUID 5SaSBe-zk7g-3XR6-5X7a-H21x-3mW4-EJyHHN--- Physical volume ---
PV Name /dev/hdc1 VG Name hxvg01 PV Size 41.93 GB / NOT usable 2.96 MB [LVM: 271.00 KB] PV# 2 PV Status available Allocatable yes (but full) Cur LV 1 PE Size (KByte) 4096 Total PE 10733 Free PE 0 Allocated PE 10733 PV UUID gQYeny-bpfF-IoQl-8dtG-867E-OXIj-FBqmlj--- Physical volume ---
PV Name /dev/hdb1 VG Name hxvg01 PV Size 76.69 GB / NOT usable 4.64 MB [LVM: 308.00 KB] PV# 3 PV Status NOT available Allocatable yes Cur LV 1 PE Size (KByte) 4096 Total PE 19631 Free PE 21 Allocated PE 19610 PV UUID 3cQksh-0OAA-ar9H-3fms-AgGQ-tOIO-eREM8v
Quoted 17 lines
vgcfgrestore -- unlocking logical volume manager
Which of the pv* vg* lv* commands are changing the physical volume (=which are dangerous for me) ?
// melle
Signature · 9 lines
+++ GMX - Mail, Messaging & more http://www.gmx.net +++ NEU: Mit GMX ins Internet. Rund um die Uhr für 1 ct/ Min. surfen! _______________________________________________ linux-lvm mailing list [email protected] http://lists.sistina.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
*** 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
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Hello,
Quoted 2 lines
Under SuSE 8.1:
- what does vgscan display ?
router:/home/melle # vgscan
vgscan -- reading all physical volumes (this may take a while...) vgscan -- ERROR "vg_read_with_pv_and_lv(): current PV" can't get data of volume group "hxvg01" from physical volume(s)
vgscan -- "/etc/lvmtab" and "/etc/lvmtab.d" successfully created vgscan -- WARNING: This program does not do a VGDA backup of your volume group
Quoted 1 line
- what does pvscan display ? Does it show your PVs on hd[cd]1 ?
router:/home/melle # pvscan
pvscan -- reading all physical volumes (this may take a while...) pvscan -- inactive PV "/dev/hdc1" is associated to unknown VG "hxvg01" (run vgscan)pvscan -- inactive PV "/dev/hdd1" is in no VG [41.93 GB] pvscan -- inactive PV "/dev/hdb1" is associated to unknown VG "hxvg01" (run vgscan)pvscan -- total: 3 [160.54 GB] / in use: 2 [118.62 GB] / in no VG: 1 [41.93 GB]
Quoted 3 lines
No need to vgcfgrestore, because your metadata should still be on hd[cd]1 unless you (or yast) write somethig onto hd[cd]1. You didn't create files systems during the update ther, did you ? ;)
uhm... no. But I created a logical volume on a new VG (I thought it's just a naming problem and tried to create VG and LV with the same names. Mounting the file system failed and I stopped any 'experimental' actions). So the filesystem should be ok, but the metadata on the PV is somehow damaged.
I guess that the job is to force the assignemt of /dev/hdd1 to VG "hxvg01" and to make this VG 'known' to the LVM and activate the PVs.
I'm a bit lost know, because I don't know how to 'write back' these information from the hxvg01.conf file to the PVs.
Any help would be deeply appreciated.
// melle
Quoted 28 lines · Thomas Mellenthin
Hello,
Quoted 2 lines
Under SuSE 8.1:
- what does vgscan display ?router:/home/melle # vgscan
vgscan -- reading all physical volumes (this may take a while...) vgscan -- ERROR "vg_read_with_pv_and_lv(): current PV" can't get data of volume group "hxvg01" from physical volume(s)
vgscan -- "/etc/lvmtab" and "/etc/lvmtab.d" successfully created vgscan -- WARNING: This program does not do a VGDA backup of your volume groupQuoted 1 line
- what does pvscan display ? Does it show your PVs on hd[cd]1 ?
router:/home/melle # pvscan
pvscan -- reading all physical volumes (this may take a while...) pvscan -- inactive PV "/dev/hdc1" is associated to unknown VG "hxvg01" (run vgscan)pvscan -- inactive PV "/dev/hdd1" is in no VG [41.93 GB] pvscan -- inactive PV "/dev/hdb1" is associated to unknown VG "hxvg01" (run vgscan)pvscan -- total: 3 [160.54 GB] / in use: 2 [118.62 GB] / in no VG: 1 [41.93 GB]Quoted 3 lines
No need to vgcfgrestore, because your metadata should still be on hd[cd]1 unless you (or yast) write somethig onto hd[cd]1. You didn't create files systems during the update ther, did you ? ;)
uhm... no. But I created a logical volume on a new VG (I thought it's just a naming problem and tried to create VG and LV with the same names. Mounting the file system failed and I stopped any 'experimental' actions). So the filesystem should be ok, but the metadata on the PV is somehow damaged.
Quoted 7 lines
I guess that the job is to force the assignemt of
/dev/hdd1to VG "hxvg01" and to make this VG 'known' to the LVM and activate the PVs.
I'm a bit lost know, because I don't know how to 'write back' these information from the hxvg01.conf file to the PVs.Any help would be deeply appreciated.
Thomas,
please send the output file named Thomas.Mellenthin.vgda.tar.bz2 of
for d in hd[b-d]1
dd if=/dev/$d of=$d.vgda bs=1k count=4k
done
tar cvf - hd[b-d]1.vgda | bzip2 > Thomas.Mellenthin.vgda.tar.bz2 rm hd[b-d]1.vgda
to me in private email ([email protected]) for further investigation.
Quoted 12 lines
// melle
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
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
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-