Spaces in LVM1 LV names => *deep trouble* when converting to LVM2
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi, I encountered a very serious bug with LVM today. I'm using lvm2-2.01.04-5 from Debian stable, kernel 2.6.17-rc3
# lvm version LVM version: 2.01.04 (2005-02-09) Library version: 1.01.00-ioctl (2005-01-17) Driver version: 4.6.0
What I did:
* created a LV in an LVM1 VG using EVMS (yeah, I know) with a space in the name (yeah, yeah, I know, I know :-));
* converted the VG to LVM2 using vgconvert;
* as a result the space in the LVM1 metadata was converted over into a space in LVM2 textual metadata, which causes a nice "Parse error line 123" for every subsequent LVM command on that VG, including "vgcfgrestore"...
Of course I did that on the root VG of my work laptop (Backups? What backups?) and ended up in the aforementioned pickle. I tried some dd stunts to overwrite the space with a dash, having discovered the correct offset and block size with strace() (which for the record were respectively one block and 2560 bytes, YMMV) - But bummer, some wiseguy put a CRC32 checksum in there! :-) I had to recompile a custom version of lvm that passes NULL as the checksum_fn parameter to text_vg_import_fd (again for the record, that's a trivial one-line patch in function _vg_read_raw_area, in format-text.c around line 286).
The situation definitely needs some fixing. My humble suggestions:
* vgconvert should surface-test the LVM1 metadata as strictly as the command-line "lvcreate" tool does on its arguments;
* there should be a mechanism for dealing with corrupt LVM metadata, at the minimum a global command-line switch to temporarily disable checksum verifications.
Now investigating backup software *real seriously* :-) Thanks for your work on the Linux LVM suite,
- --
Dominique QUATRAVAUX Ing�nieur senior
01 44 42 00 08 IDEALX
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQFEYkxmMJAKAU3mjcsRApAFAJ4tCEJ0pK1bIBWruaq5VDu2izf75gCeKQCC zzsNGFC6lPaRBvpRy2am/TM=
=ad0C
-----END PGP SIGNATURE-----
Quoted 9 lines · Dominique Quatravaux
Hi, I encountered a very serious bug with LVM today. I'm using lvm2-2.01.04-5 from Debian stable, kernel 2.6.17-rc3
# lvm version LVM version: 2.01.04 (2005-02-09) Library version: 1.01.00-ioctl (2005-01-17) Driver version: 4.6.0 What I did: * created a LV in an LVM1 VG using EVMS (yeah, I know) with a space in the name (yeah, yeah, I know, I know :-));
LVs are not permitted to have spaces in their names. EVMS should have prevented you from doing that!
Quoted 16 lines
* converted the VG to LVM2 using vgconvert;
* as a result the space in the LVM1 metadata was converted over into a space in LVM2 textual metadata, which causes a nice "Parse error line 123" for every subsequent LVM command on that VG, including "vgcfgrestore"...Of course I did that on the root VG of my work laptop (Backups? What backups?) and ended up in the aforementioned pickle. I tried some dd stunts to overwrite the space with a dash, having discovered the correct offset and block size with
strace()(which for the record were respectively one block and 2560 bytes, YMMV) - But bummer, some wiseguy put a CRC32 checksum in there! :-) I had to recompile a custom version of lvm that passes NULL as thechecksum_fnparameter totext_vg_import_fd(again for the record, that's a trivial one-line patch in function_vg_read_raw_area, informat-text.caround line 286).
Quoted 4 lines
The situation definitely needs some fixing. My humble suggestions:
* vgconvert should surface-test the LVM1 metadata as strictly as the command-line "lvcreate" tool does on its arguments;
Some metadata sanity checks are always performed before writing out metadata. They're designed to trap internal errors to stop the programs writing out invalid metadata. I'll see about enhancing them to catch this sort of problem.
Quoted 3 lines
* there should be a mechanism for dealing with corrupt LVM metadata, at the minimum a global command-line switch to temporarily disable checksum verifications.
Nope.
Alasdair
Signature · 1 line
[email protected]
Quoted 6 lines · Alasdair G Kergon
Quoted 3 lines · Dominique Quatravaux
* there should be a mechanism for dealing with corrupt LVM metadata, at the minimum a global command-line switch to temporarily disable checksum verifications.
Nope.
I should add that what you should do is extract the metadata you've retrieved off disk directly into a file, fix it in the file (differences in whitespace & comments don't matter) then use vgcfgrestore -f.
Alasdair
Signature · 1 line
[email protected]
Quoted 11 lines · Alasdair G Kergon
Quoted 9 lines · Dominique Quatravaux
Hi, I encountered a very serious bug with LVM today. I'm using lvm2-2.01.04-5 from Debian stable, kernel 2.6.17-rc3
# lvm version LVM version: 2.01.04 (2005-02-09) Library version: 1.01.00-ioctl (2005-01-17) Driver version: 4.6.0 What I did: * created a LV in an LVM1 VG using EVMS (yeah, I know) with a space in the name (yeah, yeah, I know, I know :-));
Quoted 2 lines
LVs are not permitted to have spaces in their names. EVMS should have prevented you from doing that!
Hmm...looks like the lvm2 format plugin in EVMS does check for spaces when creating a new VG or LV, but the lvm1 format plugin does not. I've added a patch to fix the lvm1 plugin. Sorry for the trouble. :( I guess this case is sufficiently unusual that no one has hit before, or at least not mentioned it to me before.
Signature · 4 lines
Kevin Corry [email protected] http://www.ibm.com/linux/ http://evms.sourceforge.net/
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Quoted 5 lines · Alasdair G Kergon
I should add that what you should do is extract the metadata you've retrieved off disk directly into a file, fix it in the file (differences in whitespace & comments don't matter) then use vgcfgrestore -f.
I tried that and it works - strange, I tried vgcfgrestore yesterday to no avail... But then I was using Ubuntu 5.10's lvm2 package as a rescue so apparently it's a different (and solved) issue. Thanks!
Knowing this, and still for the record: after any LVM operation it seems wise to try a vgchange -a y *before* rebooting to see if everything went well. Reverting a pickled LVM situation is much easier with the original toolset handy (same LVM binary and metadata backups in /etc/lvm/) than using a rescue CD!
I wrote, and you replied:
Quoted 6 lines
Quoted 3 lines
* there should be a mechanism for dealing with corrupt LVM metadata, at the minimum a global command-line switch to temporarily disable checksum verifications.
Nope.
I seem to have pushed your button, apologies for doing so. I was trying to be humorous, not disparaging: *of course* I concur that checksums are a good idea! I guess it means that humor is too slippery for us non-native speakers, oh well...
- --
Dominique QUATRAVAUX Ing�nieur senior
01 44 42 00 08 IDEALX
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQFEYu3XMJAKAU3mjcsRAugqAKC2KK88j/+w30xd6pxcn/5Juu7tuACdExXi R2b4HLcvmPmWHl+uxhbvGi0=
=tWX0
-----END PGP SIGNATURE-----