vzdump backup fails [ERROR: Device 'drive-virtio0' not found]

[philk1989]

When trying to backup a particular VM I get this error:

[COLOR=#000000]INFO: starting new backup job: vzdump 124 --remove 0 --mode snapshot --compress gzip --storage hinds-backup --node chelsea[/COLOR]
[COLOR=#000000]INFO: Starting Backup of VM 124 (qemu)[/COLOR]
[COLOR=#000000]INFO: status = running[/COLOR]
[COLOR=#000000]INFO: backup mode: snapshot[/COLOR]
[COLOR=#000000]INFO: ionice priority: 7[/COLOR]
[COLOR=#000000]INFO: creating archive '/mnt/pve/hinds-backup/dump/vzdump-qemu-124-2013_08_19-10_27_08.vma.gz'[/COLOR]
[COLOR=#000000]ERROR: Device 'drive-virtio0' not found[/COLOR]
[COLOR=#000000]INFO: aborting backup job[/COLOR]
[COLOR=#000000]ERROR: Backup of VM 124 failed - Device 'drive-virtio0' not found[/COLOR]
[COLOR=#000000]INFO: Backup job finished with errors[/COLOR]
[COLOR=#000000]TASK ERROR: job errors[/COLOR]

The log file in /var/log/vzdump/qemu-124.log shows the same thing as above.

root@chelsea:~# ls -lh /var/lib/vz/images/124/
total 23G
-rw-r--r-- 1 root root 33G Aug 19 10:31 vm-124-disk-1.qcow2

root@chelsea:~# cat /etc/pve/qemu-server/124.conf 
boot: nc
bootdisk: virtio0
cores: 1
cpu: kvm64
ide2: none,media=cdrom
memory: 4096
name: sin
net0: virtio=3A:01:9B:EE:B3:E3,bridge=vmbr0
ostype: l26
sockets: 2
virtio0: local:124/vm-124-disk-1.qcow2,size=32G

The last good backup of this VM happened on 2013-08-17. So the backup did succeed on that date. But not anytime after...

root@chelsea:/var/log/vzdump# pveversion 
pve-manager/2.3/7946f1f1

Anyone run into this before/know what is going on?

 

[dietmar]

What is the output of:


# qm monitor 124
> info block
> quit

 

[philk1989]

root@chelsea:~# qm monitor 124
Entering Qemu Monitor for VM 124 - type 'help' for help
qm> info block
drive-ide2: removable=1 locked=0 tray-open=0 io-status=ok [not inserted]
drive-sata0: removable=0 io-status=ok file=/var/lib/vz/images/124/vm-124-disk-1.qcow2 ro=0 drv=qcow2 encrypted=0 bps=0 bps_rd=0 bps_wr=0 iops=0 iops_rd=0 iops_wr=0
qm>

hmm... based on this output is it correct to assume that proxmox thinks the drive is a sata drive?

 

[dietmar]

[
hmm... based on this output is it correct to assume that proxmox thinks the drive is a sata drive?

I assume you changed the drive from sata to virtio, but forgot to restart the VM afterwards.

 

[philk1989]

I assume you changed the drive from sata to virtio, but forgot to restart the VM afterwards.

Well even if that was the case, I just restarted the VM and no luck in backing it up. The stats from qm monitor are the same.

 

[spirit]

Well even if that was the case, I just restarted the VM and no luck in backing it up. The stats from qm monitor are the same.

do you have done a stop/start ? (not reset)

 

[philk1989]

do you have done a stop/start ? (not reset)

This is what happens when I try to stop the vm.

Executing HA stop for VM 124
Local machine disabling pvevm:124...Could not connect to resource group manager
TASK ERROR: command 'clusvcadm -d pvevm:124' failed: exit code 1

 

[philk1989]

Anyone got any ideas on how to fix this?

 

[dietmar]

Seems you cluster is not correctly working. Try to read the logs in /var/log/syslog and /var/log/cluster/*

 

[philk1989]

So I found the problem... though nothing obvious in the logs that pointed to the problem. I had this particular VM setup for HA, without having HA setup correct. (My fault in misunderstanding how HA worked in proxmox.) I removed the VM from HA, activated the changes and was able to stop and start the VM. Though, I would still be curious as to why I was not able to backup the VM anymore, only because I had setup the VM for HA.

Thank you for your support in this dietmar.

Cheers,

Phil