Debian Bug report logs - #788062
os-prober corrupts LVs/partitions while being mounted inside a VM

Toggle useless messages

Message #5 received at [email protected] (full text, mbox, reply):

From: Marcel Meckel <[email protected]>

To: [email protected]

Subject: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Mon, 08 Jun 2015 10:35:42 +0200

Package: os-prober
Version: 1.65
Severity: critical

update-grub on a vhost running qemu/kvm virtual machines
while grub and os-prober is installed on the host might corrupt
guest virtual machines / guest disks.

Host
====

% lvs

  LV         VG            Attr       LSize
  root       vg_foobar_sys -wi-ao----   3.00g
  var_log    vg_foobar_sys -wi-ao---- 500.00m
  vm-luke    vg_foobar_dom -wi-ao----  40.00g

% virsh list --all

 Id    Name                    State
-------------------------------------
 9     luke                  running

% virsh domblklist luke
Target     Source
------------------------------------------------
vda        /dev/vg_foobar_dom/vm-luke

% apt-get upgrade

(linux-image-3.16.0-4-amd64 gets upgraded)
(update-grub is called)

% dmesg

Jun 08 09:31:25 sirius os-prober[7402]: debug: running 
/usr/lib/os-probes/50mounted-tests on /dev/mapper/vg_foobar_dom-vm--luke
Jun 08 09:31:25 sirius kernel: EXT4-fs (dm-3): VFS: Can't find ext4 
filesystem
Jun 08 09:31:25 sirius kernel: EXT4-fs (dm-3): VFS: Can't find ext4 
filesystem
Jun 08 09:31:25 sirius kernel: EXT4-fs (dm-3): VFS: Can't find ext4 
filesystem
Jun 08 09:31:25 sirius kernel: XFS (dm-3): Invalid superblock magic 
number
Jun 08 09:31:25 sirius kernel: FAT-fs (dm-3): utf8 is not a recommended 
IO charset for FAT filesystems, filesystem will be case sensitive!
Jun 08 09:31:25 sirius kernel: FAT-fs (dm-3): bogus number of reserved 
sectors
Jun 08 09:31:25 sirius kernel: FAT-fs (dm-3): Can't find a valid FAT 
filesystem
Jun 08 09:31:25 sirius kernel: FAT-fs (dm-3): utf8 is not a recommended 
IO charset for FAT filesystems, filesystem will be case sensitive!
Jun 08 09:31:25 sirius kernel: FAT-fs (dm-3): bogus number of reserved 
sectors
Jun 08 09:31:25 sirius kernel: FAT-fs (dm-3): Can't find a valid FAT 
filesystem
Jun 08 09:31:25 sirius kernel: VFS: Can't find a Minix filesystem V1 | 
V2 | V3 on device dm-3.
Jun 08 09:31:25 sirius kernel: hfsplus: unable to find HFS+ superblock
Jun 08 09:31:25 sirius kernel: qnx4: no qnx4 filesystem (no root dir).
Jun 08 09:31:25 sirius kernel: You didn't specify the type of your ufs 
filesystem

                               mount -t ufs -o 
ufstype=sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|nextstep-cd|openstep 
...

                               >>>WARNING<<< Wrong ufstype may corrupt 
your filesystem, default is ufstype=old



Inside guest 'luke'
===================

[1100902.892807] end_request: I/O error, dev vda, sector 10774672
[1100902.893914] Buffer I/O error on device dm-2, logical block 10002
[1100902.894781] lost page write due to I/O error on dm-2
[1100902.895219] end_request: I/O error, dev vda, sector 31670147
[1100902.897132] I/O error in filesystem ("dm-2") meta-data dev dm-2 
block 0x1400f83       ("xlog_iodone") error 5 buf count 1024
[1100902.898819] xfs_force_shutdown(dm-2,0x2) called from line 1043 of 
file 
/build/linux-2.6-pYQHNk/linux-2.6-2.6.32/debian/build/source_amd64_none/fs/xfs/xfs_log.c. 
 Return address = 0xffffffffa01fcb89
[1100902.898827] Filesystem "dm-2": Log I/O Error Detected.  Shutting 
down filesystem: dm-2
[1100902.899984] Please umount the filesystem, and rectify the 
problem(s)
[1100934.617120] Filesystem "dm-2": xfs_log_force: error 5 returned.
[1100970.617117] Filesystem "dm-2": xfs_log_force: error 5 returned.
[1101006.616137] Filesystem "dm-2": xfs_log_force: error 5 returned.


It's a bad idea to try to mount every block device found in the system
whilst the devices are being used. This could lead to catastrophic 
failure
if I/O heavy VMs get their "disks" corrupted this way.

To me this happened only to one out of 18 VMs but all but 'luke' didn't 
have much
I/O at the time.


Workaround
==========

% apt-get purge os-prober

OR

echo 'GRUB_DISABLE_OS_PROBER=true' >> /etc/default/grub

Message #10 received at [email protected] (full text, mbox, reply):

From: RedRampage <[email protected]>

To: [email protected]

Subject: Re: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Mon, 17 Aug 2015 18:12:50 +0300

I can confirm this, got the same problem with os-prober_1.63ubuntu1

Message #15 received at [email protected] (full text, mbox, reply):

From: Andrey Korolyov <[email protected]>

To: [email protected]

Subject: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Mon, 7 Sep 2015 13:24:30 +0300

The same behavior was observed for ext4 - if *any* superblock update
was made during os-prober execution, it could effectively corrupt a
filesystem in guest or, as for iSCSI, an exported one. The right
solution is to make a filesystem-specific checks like checking
EXT2_MF_BUSY before making an actual probe. Dirty and ugly patch may
land soon and I expect this issue to be re-routed to the upstream.

Message #20 received at [email protected] (full text, mbox, reply):

From: "Andreas Thalhammer (AIFB)" <[email protected]>

To: <[email protected]>

Subject: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Wed, 9 Sep 2015 11:07:34 +0200

We can confirm this Bug (Debian 8.0 Jessie on VM host and guests).

os-prober version: 1.65 amd64

VM host
=====

Sep  7 14:43:19 vm-host os-prober: debug: running 
/usr/lib/os-probes/50mounted-tests on /dev/mapper/vm1-disk

Inside VM1
=====
Sep  7 14:43:20 vm1 kernel: [4552906.712287] end_request: I/O error, dev 
vda, sector 5125812120


We had two out of five EXT4 VM filesystems corrupted in a single run of 
"apt-get upgrade". The filesystems automatically switched to read-only 
mode shortly after the first errors and we rebooted both VMs (fsck 
automatically enabled: orphaned inodes and some inodes had to be removed).

We regard this bug as highly critical.

Message #27 received at [email protected] (full text, mbox, reply):

From: Samuel Wolf <[email protected]>

To: [email protected]

Subject: Re: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Tue, 10 Nov 2015 23:25:44 +0100

[Message part 1 (text/plain, inline)]

Unfortunately I must confirm this bug which is high critical!

My workaround was apt-get --purge remove os-prober

[Message part 2 (text/html, inline)]

Message #34 received at [email protected] (full text, mbox, reply):

From: Guido Günther <[email protected]>

To: Thomas Lange <[email protected]>

Cc: [email protected], Nigel Kukard <[email protected]>

Subject: Re: tagging 788062

Date: Sun, 10 Jan 2016 18:52:04 +0100

Hi Thomas,
On Mon, Sep 28, 2015 at 02:10:11PM +0200, Thomas Lange wrote:
> tags 788062 + pending

Is there any link to a discussion how this is going to be fixed,
anything I can help with?

I discussed with Nigel (cc:) if this needs to be fixed by using
ro,noload when mounting the partition via grub-mount? The bug seems to
be there since at least

    /usr/lib/os-probes/50mounted-tests

invokes grub-mount which should use "noload"[1] for the fuse mount (IIRC).

Cheers,
 -- Guido

[1] http://www.kernel.org/doc/Documentation/filesystems/ext4.txt

Message #39 received at [email protected] (full text, mbox, reply):

From: Thomas Lange <[email protected]>

To: Guido Günther <[email protected]>

Cc: [email protected]

Subject: Re: tagging 788062

Date: Sun, 10 Jan 2016 19:34:01 +0100

>>>>> On Sun, 10 Jan 2016 18:52:04 +0100, Guido Günther <[email protected]> said:

    > Hi Thomas,
    > On Mon, Sep 28, 2015 at 02:10:11PM +0200, Thomas Lange wrote:
    >> tags 788062 + pending
I just realized that the pending tag was set by me because of an error
in my workflow. I will unset it now.

-- 
regards Thomas

Message #46 received at [email protected] (full text, mbox, reply):

From: Chris Kuehl <[email protected]>

To: [email protected]

Subject: Re: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Tue, 19 Jan 2016 22:35:53 -0800

Hi,

If this can't be fixed quickly, would it make sense to push an update
which disables os-prober on jessie in cases which are known to (even
possibly) lead to data loss?

I hate to post a "me too" message, but it's a little disheartening to
know that a bug that causes pretty bad data loss in fairly common
scenarios has been known about since June without being patched in
stable.

We just spent a lot of time and effort pinpointing and recovering from
this bug ourselves. I wish we could help others who are yet to hit this
bug, as surely more people are going to lose data until we patch this.

A simple reproduction:

    - on the host, run
        while :; do os-prober; done
    - on a guest, perform some disk IO, e.g.
        dd if=/dev/zero of=/test

On our systems, this reproduces the issue very quickly. The guest runs
under KVM and the guest's disk is an LVM logical volume. The filesystem
is ext4.

Thanks for your work on Debian!

Chris

Message #51 received at [email protected] (full text, mbox, reply):

From: Barry de Graaff <[email protected]>

To: [email protected]

Cc: Ferry Bodijn <[email protected]>, Ritchie Timmer <[email protected]>

Subject: Re: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Thu, 21 Jan 2016 11:37:54 +0100 (CET)

This bug seems to have put a few of my KVM Vm's into read-only. 

Would it be possible to make the qemu-kvm package conflict with os-prober,
not allowing both packages to be installed simultaneous.

As I don't see a use case for os-prover of a KVM host.


Thanks for considering,


Barry de Graaff


Additional logging:

Logging on the VM it looks like:

Jan 20 06:30:07 zimbra1 kernel: end_request: I/O error, dev vdb, sector 1044788760
Jan 20 06:30:07 zimbra1 kernel: end_request: I/O error, dev vdb, sector 1044788760
Jan 20 06:30:07 zimbra1 kernel: Aborting journal on device dm-1-8.
Jan 20 06:30:07 zimbra1 kernel: end_request: I/O error, dev vdb, sector 1044647936
Jan 20 06:30:07 zimbra1 kernel: Buffer I/O error on device dm-1, logical block 130580480
Jan 20 06:30:07 zimbra1 kernel: lost page write due to I/O error on dm-1
Jan 20 06:30:07 zimbra1 kernel: JBD2: I/O error detected when updating journal superblock for dm-1-8.
Jan 20 06:30:07 zimbra1 kernel: journal commit I/O error
Jan 20 06:30:07 zimbra1 kernel: EXT4-fs error (device dm-1): ext4_journal_start_sb: Detected aborted journal
Jan 20 06:30:07 zimbra1 kernel: EXT4-fs (dm-1): Remounting filesystem read-only



on the KVM host, note the time difference (osprober is fired 2 seconds before the VM goes read-only)


Jan 20 06:30:05 virtualization03 os-prober: debug: running /usr/lib/os-probes/50mounted-tests on /dev/sda1
Jan 20 06:30:05 virtualization03 kernel: [20302458.505123] XFS (sda1): Invalid superblock magic number
Jan 20 06:30:05 virtualization03 kernel: [20302458.507333] FAT-fs (sda1): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:05 virtualization03 kernel: [20302458.507470] FAT-fs (sda1): Can't find a valid FAT filesystem
Jan 20 06:30:05 virtualization03 kernel: [20302458.508497] FAT-fs (sda1): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:05 virtualization03 kernel: [20302458.508628] FAT-fs (sda1): Can't find a valid FAT filesystem
Jan 20 06:30:05 virtualization03 kernel: [20302458.509755] ntfs_warning: 4 callbacks suppressed
Jan 20 06:30:05 virtualization03 kernel: [20302458.509758] ntfs: (device sda1): is_boot_sector_ntfs(): Invalid boot sector checksum.
Jan 20 06:30:05 virtualization03 kernel: [20302458.511050] VFS: Can't find a Minix filesystem V1 | V2 | V3 on device sda1.
Jan 20 06:30:05 virtualization03 kernel: [20302458.512311] hfsplus: unable to find HFS+ superblock
Jan 20 06:30:05 virtualization03 kernel: [20302458.514701] You didn't specify the type of your ufs filesystem
Jan 20 06:30:05 virtualization03 kernel: [20302458.514701] 
Jan 20 06:30:05 virtualization03 kernel: [20302458.514701] mount -t ufs -o ufstype=sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|nextstep-cd|openstep ...
Jan 20 06:30:05 virtualization03 kernel: [20302458.514701] 
Jan 20 06:30:05 virtualization03 kernel: [20302458.514701] >>>WARNING<<< Wrong ufstype may corrupt your filesystem, default is ufstype=old
Jan 20 06:30:05 virtualization03 kernel: [20302458.514881] ufs_read_super: bad magic number
Jan 20 06:30:05 virtualization03 kernel: [20302458.517293] hfs: can't find a HFS filesystem on dev sda1
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/05efi on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 05efi: debug: Not on UEFI platform
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/10freedos on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 10freedos: debug: /dev/sda2 is not a FAT partition: exiting
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/10qnx on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 10qnx: debug: /dev/sda2 is not a QNX4 partition: exiting
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/20macosx on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 macosx-prober: debug: /dev/sda2 is not an HFS+ partition: exiting
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/20microsoft on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 20microsoft: debug: /dev/sda2 is not a MS partition: exiting
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/30utility on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 30utility: debug: /dev/sda2 is not a FAT partition: exiting
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/40lsb on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/70hurd on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/80minix on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/83haiku on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 83haiku: debug: /dev/sda2 is not a BeFS partition: exiting
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/90linux-distro on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/90solaris on mounted /dev/sda2
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/50mounted-tests on /dev/sda3
Jan 20 06:30:06 virtualization03 os-prober: debug: running /usr/lib/os-probes/50mounted-tests on /dev/sdb1
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: mounted using GRUB ext2 filesystem driver
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/05efi
Jan 20 06:30:06 virtualization03 05efi: debug: Not on UEFI platform
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/10freedos
Jan 20 06:30:06 virtualization03 10freedos: debug: /dev/sdb1 is not a FAT partition: exiting
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/10qnx
Jan 20 06:30:06 virtualization03 10qnx: debug: /dev/sdb1 is not a QNX4 partition: exiting
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/20macosx
Jan 20 06:30:06 virtualization03 macosx-prober: debug: /dev/sdb1 is not an HFS+ partition: exiting
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/20microsoft
Jan 20 06:30:06 virtualization03 20microsoft: debug: /dev/sdb1 is not a MS partition: exiting
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/30utility
Jan 20 06:30:06 virtualization03 30utility: debug: /dev/sdb1 is not a FAT partition: exiting
Jan 20 06:30:06 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/40lsb
Jan 20 06:30:07 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/70hurd
Jan 20 06:30:07 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/80minix
Jan 20 06:30:07 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/83haiku
Jan 20 06:30:07 virtualization03 83haiku: debug: /dev/sdb1 is not a BeFS partition: exiting
Jan 20 06:30:07 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/90linux-distro
Jan 20 06:30:07 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/90solaris
Jan 20 06:30:07 virtualization03 50mounted-tests: debug: running subtest /usr/lib/os-probes/mounted/efi
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/05efi on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 05efi: debug: Not on UEFI platform
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/10freedos on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 10freedos: debug: /dev/mapper/vg_root-log is not a FAT partition: exiting
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/10qnx on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 10qnx: debug: /dev/mapper/vg_root-log is not a QNX4 partition: exiting
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/20macosx on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 macosx-prober: debug: /dev/mapper/vg_root-log is not an HFS+ partition: exiting
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/20microsoft on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 20microsoft: debug: /dev/mapper/vg_root-log is not a MS partition: exiting
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/30utility on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 30utility: debug: /dev/mapper/vg_root-log is not a FAT partition: exiting
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/40lsb on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/70hurd on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/80minix on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/83haiku on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 83haiku: debug: /dev/mapper/vg_root-log is not a BeFS partition: exiting
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/90linux-distro on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/mounted/90solaris on mounted /dev/mapper/vg_root-log
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/50mounted-tests on /dev/mapper/vg_root-zimbra1--disk1
Jan 20 06:30:07 virtualization03 kernel: [20302460.135372] XFS (dm-1): Invalid superblock magic number
Jan 20 06:30:07 virtualization03 kernel: [20302460.139144] FAT-fs (dm-1): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:07 virtualization03 kernel: [20302460.139309] FAT-fs (dm-1): Can't find a valid FAT filesystem
Jan 20 06:30:07 virtualization03 kernel: [20302460.140936] FAT-fs (dm-1): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:07 virtualization03 kernel: [20302460.140982] FAT-fs (dm-1): Can't find a valid FAT filesystem
Jan 20 06:30:07 virtualization03 kernel: [20302460.142696] ntfs: (device dm-1): is_boot_sector_ntfs(): Invalid boot sector checksum.
Jan 20 06:30:07 virtualization03 kernel: [20302460.144712] VFS: Can't find a Minix filesystem V1 | V2 | V3 on device dm-1.
Jan 20 06:30:07 virtualization03 kernel: [20302460.146612] hfsplus: unable to find HFS+ superblock
Jan 20 06:30:07 virtualization03 kernel: [20302460.150001] You didn't specify the type of your ufs filesystem
Jan 20 06:30:07 virtualization03 kernel: [20302460.150001] 
Jan 20 06:30:07 virtualization03 kernel: [20302460.150001] mount -t ufs -o ufstype=sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|nextstep-cd|openstep ...
Jan 20 06:30:07 virtualization03 kernel: [20302460.150001] 
Jan 20 06:30:07 virtualization03 kernel: [20302460.150001] >>>WARNING<<< Wrong ufstype may corrupt your filesystem, default is ufstype=old
Jan 20 06:30:07 virtualization03 kernel: [20302460.150214] ufs_read_super: bad magic number
Jan 20 06:30:07 virtualization03 kernel: [20302460.153816] hfs: can't find a HFS filesystem on dev dm-1
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/50mounted-tests on /dev/mapper/vg_root-zimbra1--disk2
Jan 20 06:30:07 virtualization03 kernel: [20302460.453089] XFS (dm-2): Invalid superblock magic number
Jan 20 06:30:07 virtualization03 kernel: [20302460.457211] FAT-fs (dm-2): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:07 virtualization03 kernel: [20302460.457355] FAT-fs (dm-2): Can't find a valid FAT filesystem
Jan 20 06:30:07 virtualization03 kernel: [20302460.459210] FAT-fs (dm-2): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:07 virtualization03 kernel: [20302460.459258] FAT-fs (dm-2): Can't find a valid FAT filesystem
Jan 20 06:30:07 virtualization03 kernel: [20302460.463331] VFS: Can't find a Minix filesystem V1 | V2 | V3 on device dm-2.
Jan 20 06:30:07 virtualization03 kernel: [20302460.465399] hfsplus: unable to find HFS+ superblock
Jan 20 06:30:07 virtualization03 kernel: [20302460.469369] You didn't specify the type of your ufs filesystem
Jan 20 06:30:07 virtualization03 kernel: [20302460.469369] 
Jan 20 06:30:07 virtualization03 kernel: [20302460.469369] mount -t ufs -o ufstype=sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|nextstep-cd|openstep ...
Jan 20 06:30:07 virtualization03 kernel: [20302460.469369] 
Jan 20 06:30:07 virtualization03 kernel: [20302460.469369] >>>WARNING<<< Wrong ufstype may corrupt your filesystem, default is ufstype=old
Jan 20 06:30:07 virtualization03 kernel: [20302460.469506] ufs_read_super: bad magic number
Jan 20 06:30:07 virtualization03 kernel: [20302460.472912] hfs: can't find a HFS filesystem on dev dm-2
Jan 20 06:30:07 virtualization03 os-prober: debug: running /usr/lib/os-probes/50mounted-tests on /dev/mapper/vg_root-zimbra1--disk3
Jan 20 06:30:08 virtualization03 kernel: [20302460.986738] XFS (dm-4): Invalid superblock magic number
Jan 20 06:30:08 virtualization03 kernel: [20302461.067561] FAT-fs (dm-4): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:08 virtualization03 kernel: [20302461.067806] FAT-fs (dm-4): Can't find a valid FAT filesystem
Jan 20 06:30:08 virtualization03 kernel: [20302461.069976] FAT-fs (dm-4): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
Jan 20 06:30:08 virtualization03 kernel: [20302461.070027] FAT-fs (dm-4): Can't find a valid FAT filesystem
Jan 20 06:30:08 virtualization03 kernel: [20302461.093397] VFS: Can't find a Minix filesystem V1 | V2 | V3 on device dm-4.
Jan 20 06:30:08 virtualization03 kernel: [20302461.105483] hfsplus: unable to find HFS+ superblock
Jan 20 06:30:08 virtualization03 kernel: [20302461.109776] You didn't specify the type of your ufs filesystem
Jan 20 06:30:08 virtualization03 kernel: [20302461.109776] 
Jan 20 06:30:08 virtualization03 kernel: [20302461.109776] mount -t ufs -o ufstype=sun|sunx86|44bsd|ufs2|5xbsd|old|hp|nextstep|nextstep-cd|openstep ...
Jan 20 06:30:08 virtualization03 kernel: [20302461.109776] 
Jan 20 06:30:08 virtualization03 kernel: [20302461.109776] >>>WARNING<<< Wrong ufstype may corrupt your filesystem, default is ufstype=old
Jan 20 06:30:08 virtualization03 kernel: [20302461.109940] ufs_read_super: bad magic number
Jan 20 06:30:08 virtualization03 kernel: [20302461.113970] hfs: can't find a HFS filesystem on dev dm-4

Message #56 received at [email protected] (full text, mbox, reply):

From: Dominik George <[email protected]>

To: [email protected]

Cc: [email protected]

Subject: Re: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Tue, 02 Feb 2016 23:19:36 +0100

[Message part 1 (text/plain, inline)]

Hi,

> update-grub on a vhost running qemu/kvm virtual machines
> while grub and os-prober is installed on the host might corrupt
> guest virtual machines / guest disks.

confirmed. This just tore Teckids' main infrastructure into pieces.

-nik

-- 
PGP-Fingerprint: 3C9D 54A4 7575 C026 FB17  FD26 B79A 3C16 A0C4 F296

Dominik George · Mobil: +49-151-61623918

Teckids e.V. · FrOSCon e.V. · OpenRheinRuhr e.V.
Fellowship of the FSFE · Piratenpartei Deutschland
Opencaching Deutschland e.V. · Debian Contributor

LPIC-3 Linux Enterprise Professional (Security)

[signature.asc (application/pgp-signature, inline)]

Message #61 received at [email protected] (full text, mbox, reply):

From: Michael Lipp <[email protected]>

To: [email protected]

Date: Thu, 4 Feb 2016 16:00:25 +0100

I can confirm this bug for os-prober 1.63ubuntu2. Prober message in host 
log correlates perfectly with failure in guest log.

Message #66 received at [email protected] (full text, mbox, reply):

From: Thilo Six <[email protected]>

To: [email protected], [email protected], [email protected]

Subject: #788062 - os-prober corrupts partitions

Date: Sun, 13 Mar 2016 10:56:41 +0100

[Message part 1 (text/plain, inline)]

Hello RMs,
Hello SRMs,
Hello D-I folks,


i dare to write on this channels because there is a IMHO rather serious bug in
os-prober which up to now went below the radar of everyone it seems.
So this email in the first place is ment to make more people aware of this issue
and here is hope that someone will catch this up and fix it.

os-prober corrupts partitions
https://bugs.debian.org/788062

This bug leads to serious data loss and is TTBOMK also present in stable.
The original bugreport talks about LVM partitions being corrupted but to my
observations also simple partition layouts with extented partitions are being
hosed. Attached you will find a matching snippet from syslog where sda4 is a
extended partition.
Note the line where is says:
>>>WARNING<<< Wrong ufstype may corrupt your filesystem

IIRC i saw other messages where it talked s.th. about not valid QNX4 format.

On upgrading one machine jessie -> stretch grub asked me to choose the partition
to install so, because it said the ID/number of original partition would have
changed.

To sum it up: This one is set _valid_ to highest bug priority debian knows.


Side note:
Here i want to make crystal clear i do not blame maintainers or any other debian
folks. I am just saying there i a rather unfortunate situation which i hope will
now be handled.
Thanks for your time.


kind regards,

     Thilo

[syslog (text/plain, attachment)]

Message #79 received at [email protected] (full text, mbox, reply):

From: Colin Watson <[email protected]>

To: [email protected]

Subject: Bug#648208: fixed in os-prober 1.72

Date: Fri, 20 Jan 2017 13:04:29 +0000

Source: os-prober
Source-Version: 1.72

We believe that the bug you reported is fixed in the latest version of
os-prober, which is due to be installed in the Debian FTP archive.

A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to [email protected],
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Colin Watson <[email protected]> (supplier of updated os-prober package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing [email protected])


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Format: 1.8
Date: Fri, 20 Jan 2017 12:44:34 +0000
Source: os-prober
Binary: os-prober-udeb os-prober
Architecture: source
Version: 1.72
Distribution: unstable
Urgency: medium
Maintainer: Debian Install System Team <[email protected]>
Changed-By: Colin Watson <[email protected]>
Description:
 os-prober  - utility to detect other OSes on a set of drives
 os-prober-udeb - utility to detect other OSes on a set of drives (udeb)
Closes: 648208 674561 698598 698733 701814 776275 784709 787418 794409 801631 803155
Changes:
 os-prober (1.72) unstable; urgency=medium
 .
   * Improve logging of mounting and setting partitions to ro/rw (thanks, Ivo
     De Decker).
   * Use a read-only device-mapper entry if possible rather than setting the
     underlying device to read-only (thanks, Ivo De Decker; closes: #701814).
     Note that this introduces a dependency on dmsetup on Linux
     architectures.
   * Remove the "blockdev --setro" code path entirely, since the read-only
     device-mapper arrangement supersedes it and should be safer (closes:
     #648208).
   * Make os-prober-udeb depend on grub-mount-udeb on all Linux and kFreeBSD
     architectures, now that it's available on them all (thanks, James
     Cowgill; closes: #776275).
   * Make os-prober depend on grub-common on Linux and kFreeBSD, in order
     that grub-mount is consistently available.
   * Fix detection of /usr/ partition as a GNU/Linux root partition when
     /lib* directories are moved to /usr/ completely (thanks, Hedayat
     Vatankhah; closes: #698733).
   * Make the yaboot parser more tolerant about the syntax of "append"
     options (thanks, Hedayat Vatankhah; closes: #674561).
   * Disable debugging if OS_PROBER_DISABLE_DEBUG is set (thanks, Hedayat
     Vatankhah; closes: #698598).
   * Replace basename/dirname with shell string processing (thanks, Hedayat
     Vatankhah; part of #694668).
   * Call dmraid only once (thanks, Jeff Mahoney).
   * Fix typos in README (thanks, Nyav; closes: #803155).
   * Add os-release support (based loosely on a patch by Török Edwin; closes:
     #794409).
   * Add Devuan detection (thanks, David Hare; closes: #801631).
   * Work harder to avoid trying to mount extended partitions (thanks,
     Philippe Coval; closes: #784709).
   * Drop " (loader)" suffixes on Microsoft operating systems (thanks, Chris
     Lamb; closes: #787418).
Checksums-Sha1:
 8438248e4fbc40749b85e37c16cf433db51b3cff 1816 os-prober_1.72.dsc
 c7b3b51719328b7cf0258561f4ee9c08f9a21d79 26452 os-prober_1.72.tar.xz
Checksums-Sha256:
 f79975ddbd06ed371c3f27a781d423b2d88af6cc04896946f8df03fa42915f70 1816 os-prober_1.72.dsc
 13ed24f78e83f0c49e11635891458f067e6c8348be210adda46629dd3b7b627a 26452 os-prober_1.72.tar.xz
Files:
 90ced5506beca6dc949347ab30f6e174 1816 debian-installer optional os-prober_1.72.dsc
 16ce39ea58a684f102dd7aff1779cbf5 26452 debian-installer optional os-prober_1.72.tar.xz

-----BEGIN PGP SIGNATURE-----
Comment: Colin Watson <[email protected]> -- Debian developer
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=lgmu
-----END PGP SIGNATURE-----

Message #84 received at [email protected] (full text, mbox, reply):

From: Filippe LeMarchand <[email protected]>

To: [email protected]

Subject: Re: os-prober corrupts LVs/partitions while being mounted inside a VM

Date: Mon, 06 Feb 2017 13:03:01 +0300

In which cases does this happen? Should a vhost do high I/O at the time of os-prober running to get fs corrupted?

Send a report that this bug log contains spam.