[linux-lvm] LVM2 migration on 2.4.24 troubles - prepare for 2.6

[linux-lvm] LVM2 migration on 2.4.24 troubles - prepare for 2.6

[Axel Reinhold]

[-- Attachment #1: Type: text/plain, Size: 2813 bytes --]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

i tried to migrate my system to LVM2 on 2.4 before migration to 2.6
to avoid troubles with LVM. I got the newest device-mapper/LVM2 tools
and installed them alongside with a patched kernel and dm-mod.

When i try to "vgchange -ay" with device-mapper/LVM2 i get all of my lv's
except one! All my snapshots are also corrupt (I use snapshots intensively).

Back to LVM1 was no problem - everything running and Ok. On another
machine without snapshots moving to dm/LVM2 on 2.4.24 was no problem.

Are there incompatibilities regarding snapshots with LVM1/2?

The lvm level 9 log of vgscan/vgchange is attached (bz2 compressed).
Here comes the kernel output while activating with LVM2/dm:

$ rmmod lvm-mod
lvm -- Module successfully deactivated

$ modprobe dm-mod
device-mapper: 4.0.5-ioctl (2003-11-18) initialised: dm@uk.sistina.com

$ vgscan --mknodes
  Reading all physical volumes.  This may take a while...
  Found volume group "vg00" using metadata type lvm1
  Found volume group "vg01" using metadata type lvm1
  9 logical volume(s) in volume group "vg00" now active
  47 logical volume(s) in volume group "vg01" now active

$ vgchange -ay
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Invalid/corrupt snapshot
device-mapper: Could not create kcopyd client
device-mapper: error adding target to table
  0 logical volume(s) in volume group "vg00" now active
  15 logical volume(s) in volume group "vg01" now active

Regards
Axel Reinhold
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)

iD8DBQFACaOrHQ9JwE2bDw0RArlLAJ46zy1AxFvlwA0CxqCQmAH7HbstwgCgjsQa
ZPW1XYtXcwHsZAXdVE+o12c=
=6d3r
-----END PGP SIGNATURE-----

[-- Attachment #2: lvm.log.bz2 --]
[-- Type: application/bz2, Size: 100490 bytes --]

Re: [linux-lvm] LVM2 migration on 2.4.24 troubles - prepare for 2.6

[Alasdair G Kergon]

On Sat, Jan 17, 2004 at 10:06:11PM +0100, Axel Reinhold wrote:
> Are there incompatibilities regarding snapshots with LVM1/2?

Yes - snapshots are implemented differently: the on-disk data
is different.

Always drop snapshots before switching between LVM1 / LVM2.

Alasdair
-- 
agk@uk.sistina.com

Re: [linux-lvm] LVM2 migration on 2.4.24 troubles - prepare for 2.6

[Axel Reinhold]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

According to Alasdair G Kergon:
> On Sat, Jan 17, 2004 at 10:06:11PM +0100, Axel Reinhold wrote:
> > Are there incompatibilities regarding snapshots with LVM1/2?
> 
> Yes - snapshots are implemented differently: the on-disk data
> is different.
> 
> Always drop snapshots before switching between LVM1 / LVM2.
> 

Dear Alasdair,

thanks for your answer.  I use snapshots intensively.  I
currently have 6 monthly, 5 weekly and 7 daily snapshots active
with a rotating algorithm.

Does the new implementation copy dirty blocks also in "all"
active snapshots or is it now possible to use the cow-blocks
"cross-snapshot"?  I mean that a new dirty block is only copied
into the oldest snapshot and all the other ones can use that one.
NetApp's Filer does it this way.

Regards
Axel Reinhold

- -- 
|------------------------+--------------------------------------|
| Axel Reinhold          | Fax:   +49-9287-8244                 |
| Franz-Heinrich-Str. 20 | eMail: axel@freakout.de              |
| 95100 Selb             | http://www.freakout.de               |
| Germany          | Please do not send more than 100 kilobytes |
|------------------+--------------------------------------------|
| Fingerprint: 8D EF 9F 22 DF 9A 9B 68  E5 8C 12 C7 8D 6A 97 4E |
|---------------------------------------------------------------|
| Legal Warning: Do NOT send unsolicited commercial email to me |

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 5.0i for non-commercial use
Charset: noconv

iQA/AwUBQA135R0PScBNmw8NEQLNzgCgsimp2AZxvlenB0Wc0OeMf5P/OEwAn1xc
fBfc6W6eaWeOKMdLLTGlxf/T
=0pdu
-----END PGP SIGNATURE-----