Ganeti Deboostrap Trusty - ALERT! /dev/vda1 does not exist. Dropping to a shell

[Jon Schipp]

Hello all, I'm receiving the following error using SUITE=trusty for debootstrap
and am looking for advice on how to get the VM to boot.

ALERT! /dev/vda1 does not exist. Dropping to a shell!

root@kvm1:~# gnt-cluster version
Software version: 2.9.3
Internode protocol: 2090000
Configuration format: 2090000
OS api version: 20
Export interface: 0
VCS version: v2.9.3

root@kvm1:~# lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 14.04.1 LTS
Release: 14.04
Codename: trusty

root@kvm1:~# uname -a
Linux kvm1.local 3.13.0-44-generic #73-Ubuntu SMP Tue Dec 16 00:22:43 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux

root@kvm1:~# kvm --version
QEMU emulator version 2.0.0 (Debian 2.0.0+dfsg-2ubuntu1.10), Copyright (c) 2003-2008 Fabrice Bellard

root@kvm1:~# dpkg -l | grep debootstrap
ii debootstrap 1.0.59ubuntu0.2 all Bootstrap a basic Debian system
ii ganeti-instance-debootstrap 0.11-1 all debootstrap-based instance OS definition for ganeti

root@kvm1:~# grep EXTRA /etc/ganeti/instance-debootstrap/variants/default.conf
EXTRA_PKGS="sudo,vim,openssh-server,grub-pc"

root@kvm1:~# gnt-instance add -t drbd -o debootstrap+default -n kvm1.local:kvm2.local -s 4g instance3
Fri Feb 6 00:38:45 2015 * creating instance disks...
Fri Feb 6 00:38:53 2015 adding instance instance3 to cluster config
Fri Feb 6 00:39:00 2015 - INFO: Waiting for instance instance3 to sync disks
Fri Feb 6 00:39:11 2015 - INFO: - device disk/0: 29.00% done, 44s remaining (estimated)
Fri Feb 6 00:39:56 2015 - INFO: Instance instance3's disks are in sync
Fri Feb 6 00:39:56 2015 * running the instance OS create scripts...
Failure: command execution error:
Could not add os for instance instance3 on node kvm1.local: OS create script failed (exited with exit code 1), last lines in the log file:
Need to get 924 kB of archives.
After this operation, 1090 kB of additional disk space will be used.
WARNING: The following packages cannot be authenticated!

grub-legacy

Get:1 http://ftp.us.debian.org/debian/ wheezy/main grub-legacy amd64 0.97-67 [924 kB]
Fetched 924 kB in 0s (997 kB/s)
(Reading database ... 11873 files and directories currently installed.)
Removing grub-pc ...
Removing grub2-common ...
Processing triggers for man-db ...
Processing triggers for install-info ...
Selecting previously unselected package grub-legacy.
(Reading database ... 11845 files and directories currently installed.)
Unpacking grub-legacy (from .../grub-legacy_0.97-67_amd64.deb) ...
Processing triggers for man-db ...
Setting up grub-legacy (0.97-67) ...
Searching for GRUB installation directory ... found: /boot/grub
warning: grub-probe can't find drive for /dev/sda1.
grub-probe: error: cannot find a GRUB drive for /dev/sda1. Check your device.map.
run-parts: /etc/ganeti/instance-debootstrap/hooks/grub exited with return code 1

root@kvm1:~# gnt-instance shutdown --timeout=0 instance3
Waiting for job 506 for instance3 ...

root@kvm1:~# gnt-instance start -H kernel_path=/boot/vmlinuz-3.13.0-44-generic,initrd_path=/boot/initrd.img-3.13.0-44-generic,kernel_args="ro single" instance3
Waiting for job 507 for instance3 ...

root@kvm1:~# gnt-instance console instance3
[ 31.094613] random: lvm urandom read with 29 bits of entropy available
Gave up waiting for root device. Common problems:
- Boot args (cat /proc/cmdline)
- Check rootdelay= (did the system wait long enough?)
- Check root= (did the system wait for the right device?)
- Missing modules (cat /proc/modules; ls /dev)
ALERT! /dev/vda1 does not exist. Dropping to a shell!

BusyBox v1.21.1 (Ubuntu 1:1.21.0-1ubuntu1) built-in shell (ash)
Enter 'help' for a list of built-in commands.
(initramfs)

/dev/vda1 doesn't exist in /dev
root@instance3:/dev# ls -l vda*
brw-rw---- 1 root disk 253, 0 Feb 6 05:41 vda

root@instance2:/dev# fdisk -l

Disk /dev/vda: 4294 MB, 4294967296 bytes
16 heads, 63 sectors/track, 8322 cylinders, total 8388608 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/vda doesn't contain a valid partition table

Also, tried this after getting it to boot from initramfs by symlinking /dev/vda1 to /dev/vda
(initramfs) cd /dev
(initramfs) ln -s vda1 vda
(initramfs)
Begin: Running /scripts/local-premount ... done.
...
Stopping Send an event to indicate plymouth is up [ OK ]
The disk drive for / is not ready yet or not present.
keys:Continue to wait, or Press S to skip mounting or M for manual recovery
* Starting Mount filesystems on boot [fail]
* Starting set console keymap [ OK ]
Root filesystem check failed.
A maintenance shell will now be started.
CONTROL-D will terminate this shell and reboot the system.
Press enter for maintenance(or type Control-D to continue):

root@instance3:~# grub-install /dev/vda
df: Warning: cannot read table of mounted file systems: No such file or directory
df: Warning: cannot read table of mounted file systems: No such file or directory

Any help is appreciated, thanks.

Thanks
--
Jon Schipp,
jonschipp.com, sickbits.net, opennsm.ncsa.illinois.edu

[Jon Schipp]

I have tried the 2 grub hooks mentioned here [1] for Precise which use grub2 (grub-pc) with no luck.
I removed the grub packages from EXTRA_PKGS which still results in an error but there's less of them.
Note: that I plan to use drbd guests only. See script I have installed as hooks/grub currently [2].

$ cat /etc/ganeti/instance-debootstrap/variants/default.conf 
ARCH="amd64"
EXTRA_PKGS="sudo,vim,openssh-server"

$ gnt-instance add -t drbd -o debootstrap+default -n kvm1.local:kvm2.local -s 4g instance2

Tue Feb 10 19:32:38 2015 * creating instance disks...
Tue Feb 10 19:32:45 2015 adding instance instance2 to cluster config
Tue Feb 10 19:32:51 2015  - INFO: Waiting for instance instance2 to sync disks
Tue Feb 10 19:32:59 2015  - INFO: - device disk/0: 20.70% done, 50s remaining (estimated)
Tue Feb 10 19:33:49 2015  - INFO: Instance instance2's disks are in sync
Tue Feb 10 19:33:49 2015 * running the instance OS create scripts...

Failure: command execution error:
Could not add os for instance instance2 on node kvm1.local: OS create script failed (exited with exit code 1), last lines in the log file:
Setting up os-prober (1.63ubuntu1) ...
Processing triggers for ureadahead (0.100.0-16) ...
Setting up grub2-common (2.02~beta2-9) ...
Setting up grub-pc-bin (2.02~beta2-9) ...
Setting up grub-gfxpayload-lists (0.6) ...
Setting up grub-pc (2.02~beta2-9) ...

Creating config file /etc/default/grub with new version
grub-probe: error: cannot find a device for / (is /dev mounted?).
grub-probe: error: cannot find a device for /boot (is /dev mounted?).
grub-probe: error: cannot find a device for /boot/grub (is /dev mounted?).
Processing triggers for libc-bin (2.19-0ubuntu6) ...

Generating grub configuration file ...
/proc/devices: fopen failed: No such file or directory
device node not found
device node not found
/proc/devices: fopen failed: No such file or directory
device node not found
grub-probe: error: cannot find a GRUB drive for /dev/vda1.  Check your device.map.

run-parts: /etc/ganeti/instance-debootstrap/hooks/grub exited with return code 1

Is there a known working grub hook for Ubuntu Trusty?

Any help is greatly appreciated, I've been working on this on and off for a few weeks now in my spare time for a local Linux User Group.

[1] https://code.google.com/p/ganeti/issues/detail?id=549
[2] https://code.google.com/p/ganeti/issues/attachmentText?id=549&aid=5490005000&name=grub&token=ABZ6GAfvirzN9hpDz4Ebc9Q3dAUSxvHveQ%3A1423616124704

[ge...@riseup.net]

Jon,

On 15-02-10 17:44:35, Jon Schipp wrote:
> [...]

> Is there a known working grub hook for Ubuntu Trusty?

> [...]

If you're okay with extlinux instead of grub, consider the following
code. I'm using this without any problems on Debian wheezy / jessie,
don't know if it works with Ubuntu aswel, but I guess so:

----

set -e

. common.sh

CLEANUP=( )
set -x
trap cleanup EXIT

if [ -z "$TARGET" -o ! -d "$TARGET" ]; then
echo "Missing target directory"
exit 1
fi

# allow extlinux to find device
mount --bind /dev $TARGET/dev
CLEANUP+=("umount $TARGET/dev")
mount --bind /proc $TARGET/proc
CLEANUP+=("umount $TARGET/proc")

# generate configuration
echo 'EXTLINUX_PARAMETERS="ro console=ttyS0,38400n8"' > $TARGET/etc/default/extlinux
chroot "$TARGET" extlinux-update

# install extlinux
chroot "$TARGET" extlinux -i /boot/extlinux

# install boot record
dd if="$TARGET/usr/lib/extlinux/mbr.bin" of=$BLOCKDEV

cleanup
trap - EXIT

exit 0

---

HTH,
Georg

[candlerb]

On Saturday, 7 February 2015 23:28:02 UTC, Jon Schipp wrote:

Fri Feb  6 00:39:56 2015 * running the instance OS create scripts...
Failure: command execution error:
Could not add os for instance instance3 on node kvm1.local: OS create script failed (exited with exit code 1), last lines in the log file:

...

Searching for GRUB installation directory ... found: /boot/grub
warning: grub-probe can't find drive for /dev/sda1.
grub-probe: error: cannot find a GRUB drive for /dev/sda1.  Check your device.map.
run-parts: /etc/ganeti/instance-debootstrap/hooks/grub exited with return code 1

So clearly grub didn't install, and the usual solution would be to start from a host kernel and run grub manually inside the VM to fix it.

 

root@kvm1:~# gnt-instance start -H kernel_path=/boot/vmlinuz-3.13.0-44-generic,initrd_path=/boot/initrd.img-3.13.0-44-generic,kernel_args="ro single"  instance3
Waiting for job 507 for instance3 ...

root@kvm1:~# gnt-instance console instance3
[   31.094613] random: lvm urandom read with 29 bits of entropy available
Gave up waiting for root device.  Common problems:
 - Boot args (cat /proc/cmdline)
   - Check rootdelay= (did the system wait long enough?)
   - Check root= (did the system wait for the right device?)
 - Missing modules (cat /proc/modules; ls /dev)
ALERT!  /dev/vda1 does not exist.  Dropping to a shell!

Ugh, kernel boots but can't find root filesystem.

 

BusyBox v1.21.1 (Ubuntu 1:1.21.0-1ubuntu1) built-in shell (ash)
Enter 'help' for a list of built-in commands.
(initramfs)

/dev/vda1 doesn't exist in /dev
root@instance3:/dev# ls -l vda*
brw-rw---- 1 root disk 253, 0 Feb  6 05:41 vda

root@instance2:/dev# fdisk -l

Disk /dev/vda: 4294 MB, 4294967296 bytes
16 heads, 63 sectors/track, 8322 cylinders, total 8388608 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/vda doesn't contain a valid partition table

Ugh, no partition table!

What version of ganeti-instance-debootstrap, and how did you configure it?

Did you configure it to install without a partition table, i.e. so /dev/vda is the whole filesystem? (To test this, you should be able to find the ganeti device node and mount it directly. Given that you're using drbd, you should activate the disks first. To be honest, it would be simpler to start with -t plain and then when you have the problem cracked move onto -t drbd)

But the default today should be to use a partition table. From /etc/default/ganeti-instance-debootstrap:

# PARTITION_STYLE: whether and how the target device should be partitioned.

# Allowed values:

# 'none': just format the device, but don't partition it

# 'msdos': install an msdos partition table on the device, with a single

#          partition on it

# (more styles may be added in the future)

# The default is "msdos" from ganeti 2.0 onwards, but none if installing under

# Ganeti 1.2 (os api version 5)

# PARTITION_STYLE="none"

[Jon Schipp]

Thanks for the responses.

I'm using ganeti-instance-debootstrap 0.14-2 now from https://launchpad.net/ubuntu/+source/ganeti-instance-debootstrap/0.14-2/+build/6036863

$ grep -v ^# /etc/default/ganeti-instance-debootstrap
ARCH=""
SUITE="trusty"
EXTRA_PKGS=""
CUSTOMIZE_DIR="/etc/ganeti/instance-debootstrap/hooks"
GENERATE_CACHE="yes"
CLEAN_CACHE="14"
PARTITION_STYLE="msdos"

$ cat /etc/ganeti/instance-debootstrap/variants/trusty.conf 
MIRROR="http://archive.ubuntu.com/ubuntu/"
SUITE="trusty"
EXTRA_PKGS="openssh-server,htop"
COMPONENTS="main,restricted,universe,multiverse"
ARCH="amd64"

Did you configure it to install without a partition table, i.e. so /dev/vda is the whole filesystem? (To test this, you should be able to find the ganeti device node and mount it directly. Given that you're using drbd, you should activate the disks first. To be honest, it would be simpler to start with -t plain and then when you have the problem cracked move onto -t drbd)

Where are the ganeti device nodes?
 

But the default today should be to use a partition table. From /etc/default/ganeti-instance-debootstrap:

# PARTITION_STYLE: whether and how the target device should be partitioned.

# Allowed values:

# 'none': just format the device, but don't partition it

# 'msdos': install an msdos partition table on the device, with a single

#          partition on it

# (more styles may be added in the future)

# The default is "msdos" from ganeti 2.0 onwards, but none if installing under

# Ganeti 1.2 (os api version 5)

# PARTITION_STYLE="none"

I've tried both none and msdos with the same results.

With plain, added set -x to the grub hook:

$ gnt-instance add -t plain -o debootstrap+default -s 4g instance2

Could not add os for instance instance2 on node kvm2.local: OS create script failed (exited with exit code 1), last lines in the log file:
+ chroot /tmp/tmp.CBlfxXFrvn update-grub
Generating grub.cfg ...
cat: /boot/grub/video.lst: No such file or directory
grep: /proc/mounts: No such file or directory
Cannot find list of partitions!  (Try mounting /sys.)
done
+ grub-install --no-floppy --root-directory=/tmp/tmp.CBlfxXFrvn /tmp/tmp.CBlfxXFrvn/dev/vda
Installing for i386-pc platform.
grub-install: error: disk `lvmid/25U3El-4gv1-3RLJ-pfZd-Uptq-pxQ5-SxAbb4/9q8W78-gofx-qvhS-TSYV-IGi5-rjRo-4N7NNp' not found.
+ cleanup
+ '[' 2 -gt 0 ']'
+ LAST_ELEMENT=1
++ seq 1 -1 0
+ REVERSE_INDEXES='1
0'
+ for i in '$REVERSE_INDEXES'
+ rm -f /tmp/tmp.CBlfxXFrvn/dev/vda1
+ for i in '$REVERSE_INDEXES'
+ rm -f /tmp/tmp.CBlfxXFrvn/dev/vda

[Jon Schipp]

Georg,

I placed this in hooks/extlinux and added a new instance, the instance is added almost instantly
and then CPU maxes out. It doesn't make it to the scripts (see log)

$ gnt-instance add -t plain -o debootstrap+default -n kvm1.local -s 4g instance3

Wed Feb 11 17:29:01 2015 * disk 0, size 4.0G
Wed Feb 11 17:29:01 2015 * creating instance disks...
Wed Feb 11 17:29:02 2015 adding instance instance3 to cluster config
Wed Feb 11 17:29:03 2015  - INFO: Waiting for instance instance3 to sync disks
Wed Feb 11 17:29:03 2015  - INFO: Instance instance3's disks are in sync
Wed Feb 11 17:29:03 2015 * running the instance OS create scripts...
Wed Feb 11 17:29:09 2015 * starting instance...

log file:

BLKRRPART: Invalid argument
Disk /dev/xenvg/3313e707-7a9b-430a-8029-e748afe9b6c7.disk0: 4096 cylinders, 64 heads, 32 sectors/track
sfdisk: ERROR: sector 0 does not have an msdos signature
 /dev/xenvg/3313e707-7a9b-430a-8029-e748afe9b6c7.disk0: unrecognized partition table type

Old situation:
No partitions found

New situation:
Units = sectors of 512 bytes, counting from 0

   Device Boot    Start       End   #sectors  Id  System
/dev/xenvg/3313e707-7a9b-430a-8029-e748afe9b6c7.disk0p1   *      2048   8388607    8386560  83  Linux
/dev/xenvg/3313e707-7a9b-430a-8029-e748afe9b6c7.disk0p2             0         -          0   0  Empty
/dev/xenvg/3313e707-7a9b-430a-8029-e748afe9b6c7.disk0p3             0         -          0   0  Empty
/dev/xenvg/3313e707-7a9b-430a-8029-e748afe9b6c7.disk0p4             0         -          0   0  Empty

Successfully wrote the new partition table
Re-reading the partition table ...

BLKRRPART: Invalid argument
passwd: password expiry information changed.

[Jon Schipp]

Here's the detailed log for debootstrap add for
$ gnt-instance add -t plain -o debootstrap+default -n kvm1.local -s 4g instance2

https://gist.github.com/jonschipp/06cd9dc3f1d54c4df324#file-ganeti-trusty-grub-error

[candlerb]

On Wednesday, 11 February 2015 23:24:58 UTC, Jon Schipp wrote:

Where are the ganeti device nodes?

# gnt-node volumes

will list all the instances and their LVM volumes.

For a "plain" instance you can access

/dev/<volgroup>/<uuid>.disk0

But *never* access this for a drbd instance. Instead use gnt-instance activate-disks, following the instructions at

http://docs.ganeti.org/ganeti/master/html/admin.html#accessing-an-instance-s-disks

 

So you can try fdisk or kpartx on this node to see if it's really partitioned or not, and there's no need to start any VM. 

If you're not getting a partition table then you need to sort this out first, before you look at any grub hooks.

Note also: when you do get to the point of using the grub hook, last time I tried it was broken:

https://code.google.com/p/ganeti/issues/detail?id=549

There was a version added to this ticket which worked with -t plain but not with -t drbd. I see that now someone has added a version which is supposed to work with both.

However I've long since moved away from debootstrap, and am using snf-image instead. Since snf-image unpacks an existing image containing partition table and boot loader, this sort of issue simply doesn't happen.

Regards,

Brian.

[Jon Schipp]

Thanks for the help, I think I understand how this works a bit better now.
I have plain working now.

On Thu, Feb 12, 2015 at 2:37 AM, candlerb <b.ca...@pobox.com> wrote:

On Wednesday, 11 February 2015 23:24:58 UTC, Jon Schipp wrote:

Where are the ganeti device nodes?

# gnt-node volumes

will list all the instances and their LVM volumes.

For a "plain" instance you can access

/dev/<volgroup>/<uuid>.disk0

But *never* access this for a drbd instance. Instead use gnt-instance activate-disks, following the instructions at

http://docs.ganeti.org/ganeti/master/html/admin.html#accessing-an-instance-s-disks

 

So you can try fdisk or kpartx on this node to see if it's really partitioned or not, and there's no need to start any VM. 

If you're not getting a partition table then you need to sort this out first, before you look at any grub hooks.

There is a partition table :)

$ fdisk -l /dev/xenvg/8464a193-5e24-4a40-8909-7a35ec607f30.disk0 /mnt/

Disk /dev/xenvg/8464a193-5e24-4a40-8909-7a35ec607f30.disk0: 4294 MB, 4294967296 bytes
64 heads, 32 sectors/track, 4096 cylinders, total 8388608 sectors

Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

                                                 Device Boot      Start         End      Blocks   Id  System
/dev/xenvg/8464a193-5e24-4a40-8909-7a35ec607f30.disk0p1   *        2048     8388607     4193280   83  Linux 

Note also: when you do get to the point of using the grub hook, last time I tried it was broken:

https://code.google.com/p/ganeti/issues/detail?id=549

I went through this thread earlier this week and posted a follow up message here.
I tried both of them and it's not working for me unfortunately.

 

There was a version added to this ticket which worked with -t plain but not with -t drbd. I see that now someone has added a version which is supposed to work with both.

However I've long since moved away from debootstrap, and am using snf-image instead. Since snf-image unpacks an existing image containing partition table and boot loader, this sort of issue simply doesn't happen.

Sweet, I'll have to try this out.

Booting into the instance using my host's kernel showed me that device.map doesn't look right and no kernels were installed.

root@instance2:~# cat /boot/grub/device.map
(hd0) /tmp/tmp.Q8kqtHw8tj/dev/vda 

root@instance2:~# ls -l /boot/
total 4
drwxr-xr-x 5 root 114 4096 Feb 13 00:32 grub

I'm able to fix this by

root@instance2:~# apt-get install linux-image-generic
root@instance2:~# cat /boot/grub/device.map
(hd0) /dev/vda
root@instance2:~# update-grub
root@instance2:~# grub-install --no-floppy /dev/vda
Installation finished. No error reported.
...

Ubuntu 14.04 LTS instance2 ttyS0

instance2 login: 

Though, chroot "$TARGET"  apt-get -y --force-yes install linux-image-generic grub-pc in hooks/grub fails
so I'm doing it manually for now. /proc is not available at that state.

Selecting previously unselected package linux-image-3.13.0-24-generic.
Preparing to unpack .../linux-image-3.13.0-24-generic_3.13.0-24.46_amd64.deb ...
grep: /proc/cpuinfo: No such file or directory
This kernel does not support a non-PAE CPU.
dpkg: error processing archive /var/cache/apt/archives/linux-image-3.13.0-24-generic_3.13.0-24.46_amd64.deb (--unpack):
subprocess new pre-installation script returned error exit status 1

Thanks

[ge...@riseup.net]

On 15-02-12 19:25:56, Jon Schipp wrote:
> Thanks for the help, I think I understand how this works a bit better now.
> I have plain working now.

For people who maybe run into similar problems: Could you please
describe what you did to solve the problem?

[ge...@riseup.net]

On 15-02-11 17:56:09, Jon Schipp wrote:
> I placed this in hooks/extlinux and added a new instance, the instance is
> added almost instantly
> and then CPU maxes out. It doesn't make it to the scripts (see log)
>
> $ gnt-instance add -t plain -o debootstrap+default -n kvm1.local -s 4g
> instance3
>
> Wed Feb 11 17:29:01 2015 * disk 0, size 4.0G
> Wed Feb 11 17:29:01 2015 * creating instance disks...
> Wed Feb 11 17:29:02 2015 adding instance instance3 to cluster config
> Wed Feb 11 17:29:03 2015 - INFO: Waiting for instance instance3 to sync
> disks
> Wed Feb 11 17:29:03 2015 - INFO: Instance instance3's disks are in sync
> Wed Feb 11 17:29:03 2015 * running the instance OS create scripts...
> Wed Feb 11 17:29:09 2015 * starting instance...

Just for the record: It made it to the scripts, passed these
successfully (see "running the instance OS create scripts...", which is
different to the situation you described before using your grub-hook)
and afterwards started the instance, which wasn't successfull, thus the
maxed out cpu.

[Jon Schipp]

Georg, I responded inline,  the problem and solution were at the bottom of my previous e-mail.

[ge...@riseup.net]

On 2015-02-13 17:58, Jon Schipp wrote:
> Georg, I responded inline, the problem and solution were at the bottom
> of
> my previous e-mail.

Sorry Jon, I've overlooked this.

[Jon Schipp]

No worries. I'm still having trouble getting the grub hook to work.
It fails on ``chroot "$TARGET" update-grub'' with "grub-probe: error: failed to get canonical path of..." 
Does anyone know how to solve this?

Setting up grub2-common (2.02~beta2-9) ...
Setting up grub-pc-bin (2.02~beta2-9) ...
Setting up grub-gfxpayload-lists (0.6) ...
Setting up grub-pc (2.02~beta2-9) ...

Creating config file /etc/default/grub with new version

grub-probe: error: failed to get canonical path of `/dev/mapper/xenvg-ba3ee6b1--5643--4e74--838d--52c0e1402347.disk0-1'.
grub-probe: error: failed to get canonical path of `/dev/mapper/xenvg-ba3ee6b1--5643--4e74--838d--52c0e1402347.disk0-1'.
grub-probe: error: failed to get canonical path of `/dev/mapper/xenvg-ba3ee6b1--5643--4e74--838d--52c0e1402347.disk0-1'.

Processing triggers for libc-bin (2.19-0ubuntu6) ...

+ CLEANUP+=("umount $TARGET/proc")
+ CLEANUP+=("umount $TARGET/sys")
++ stat -L -c '0x%t 0x%T' /dev/xenvg/ba3ee6b1-5643-4e74-838d-52c0e1402347.disk0
+ mknod /tmp/tmp.OsrzuBUhBm/dev/vda b 0xfc 0x3
+ CLEANUP+=("rm -f $TARGET/dev/vda")
++ stat -L -c '0x%t 0x%T' /dev/mapper/xenvg-ba3ee6b1--5643--4e74--838d--52c0e1402347.disk0-1
+ mknod /tmp/tmp.OsrzuBUhBm/dev/vda1 b 0xfc 0x4
+ CLEANUP+=("rm -f $TARGET/dev/vda1")
+ chroot /tmp/tmp.OsrzuBUhBm update-grub
/usr/sbin/grub-probe: error: failed to get canonical path of `/dev/mapper/xenvg-ba3ee6b1--5643--4e74--838d--52c0e1402347.disk0-1'.
...

run-parts: /etc/ganeti/instance-debootstrap/hooks/grub exited with return code 1