[clug] LVM recovery -- Canberra LUG a last resort!

[Adam Thomas]

Hi Daniel,

Have you tried any of the disk recovery tools? I've had good
experience with TestDisk[0]. Your problem looks more complex than mine
was but TestDisk has support for LVM and Linux RAID MD so it's
probably worth a try.

Are you able to 'dd' the disk into an image backup? That way you still
have recovery options if you completely mangle the copy of the data.

HTH

Adam.

[0] http://www.cgsecurity.org/wiki/TestDisk

On 04/11/2007, Daniel Rose <drose at nla.gov.au> wrote:
> Hi everyone,
>
> I'd really really like some help with LVM, and it's probably not very hard.
>
> I will happily pay for a good result, with cash, physical labour,
> charitable donations, catering for the next meeting or anything else you
> like.  I'm in Belconnen.
>
> I've tried people on IRC, 2 web forums, two well-respected peers and the
> man pages, but I'm really, really stuck.
>
> I have the hard drive from an SME server (aka e-smith, at contribs.org)
> which I wish to read data from.
>
> The hardware is fine.  There are 2 partitions, /boot and an LVM
> partition which has root and swap partitions 'inside'.
>
> I found that my other box couldn't read the LVM software raid partition,
> probably because the kernel had no md support.  That's fixed now, but in
> trying various commands I have changed something which should not have
> changed, and the drive no longer boots by itself (can't find the raid
> set 'main')
>
> All I think I've altered on disk are these bytes near the beginning of
> the partition.
>
> [...binary...]LABELONE  [...binary...]  LVM2
> 0016oVVgdWIVKqB0yPXk2Acu5ZxcIgMj5JD [...binary...]
>
> By doing
> pvcreate -u 6oVVgd-WIVK-qB0y-PXk2-Acu5-ZxcI-gMj5JD /dev/hdd2
> Software RAID md superblock detected on /dev/hdd2. Wipe it? [y/n] n
> Physical volume "/dev/hdd2$B!m(B successfully created
>
> As suggested by someone somewhere on the Internet.
>
> After this, a pvdisplay shows good info, but a second run of pvdisplay
> shows nothing.  Something strange happens during the execution of
> pvdisplay so that subsequent runs of the command find nothing.
>
> The data on disk does not change between good and bad runs of pvdisplay;
> at least, not the first 10kbytes.
>
> Normally I think after the pvcreate succeeds, one should be able to make
> a volume group, however:
>
>  vgcreate -v main  /dev/hdd2
>     Wiping cache of LVM-capable devices
>     Adding physical volume '/dev/hdd2' to volume group 'main'
>   /dev/hdd2 not identified as an existing physical volume
>   Unable to add physical volume '/dev/hdd2' to volume group 'main'.
>
> Which is a shame.
>
>
> So... I'm not sure what's going on.  Further along the disk can be found
> text files detailing the construction, first with seqno1, 2 and finally
> this (edited) promise of potential success:
>
> main {
> id = "em7z2K-GhvB-mpW4-YNjE-SiZ7-zia2-PIObh2"
> seqno = 3
> status = ["RESIZEABLE", "READ", "WRITE"]
> extent_size = 65536
> max_lv = 0
> max_pv = 0
> physical_volumes {    pv0 {    id = "6oVVgd-WIVK-qB0y-PXk2-Acu5-ZxcI-gMj5JD"
> device = "/dev/md2"
> status = ["ALLOCATABLE"]
> pe_start = 384
> pe_count = 2381}}
>
> logical_volumes {   root {id = "Cu9udo-QFcL-72XW-G1OU-1sQA-giBa-WASzs2"
> status = ["READ", "WRITE", "VISIBLE"]
> segment_count = 1
> segment1 { start_extent = 0 extent_count = 2372 type = "striped"
> stripe_count = 1        # linear
> stripes = [ "pv0", 0]}}
>
> swap {id = "aM7Jm4-QfMW-4y2c-zCv3-0O6f-CVzz-P3Ebg4"
> status = ["READ", "WRITE", "VISIBLE"]
> segment_count = 1
> segment1 {start_extent = 0 extent_count = 8 type = "striped"
> stripe_count = 1        # linear
> stripes = [ "pv0", 2372 ]}}}}
> # Generated by LVM2: Tue Nov 21 18:40:21 2006
>
>
> It seems to me that given this information, it should not be
> particularly difficult to do this.  I'm impatient though, and I don't
> have a ready 80Gb to dump a copy to, and I'm stressed that I'll break
> something important if I keep trying different things.
>
> I'd be delighted if anyone is interested in offering any advice at all;
> I can even enable remote ssh for anyone feeling particularly generous
> with their time.
>
>
> Regards,
>
> Daniel Rose
> --
> linux mailing list
> linux at lists.samba.org
> https://lists.samba.org/mailman/listinfo/linux
>