migration fail. Seems like bug.

[mir]

Error message:
task started by HA resource agent
Jan 01 00:22:51 starting migration of VM 117 to node 'esx1' (192.168.2.8)
Jan 01 00:22:51 copying disk images
Jan 01 00:22:51 ERROR: Failed to sync data - can't do online migration - VM uses local disks
Jan 01 00:22:51 aborting phase 1 - cleanup resources
Jan 01 00:22:51 ERROR: migration aborted (duration 00:00:00): Failed to sync data - can't do online migration - VM uses local disks
TASK ERROR: migration aborted

cat /etc/pve/qemu-server/117.conf
#eth0%3A 172.16.1.7
bootdisk: virtio0
cores: 1
cpu: qemu32
memory: 512
name: proxy
net0: virtio=1A:1D:E8:65:38:8C,bridge=vmbr1
ostype: l26
sockets: 1
virtio0: qnap_lvm:vm-117-disk-1,cache=writeback,size=4G

How can I solve this?

 

[mir]

I have removed it from HA and even rebooted the node but the error message I get is the same.

 

[udo]

I have removed it from HA and even rebooted the node but the error message I get is the same.

Hi,
are there unused disks (which aren't in the config)?

Udo

 

[mir]

I, sort of, fixed it:-\

Problem:
For reasons unknown to me /var/lib/vz/images contained:
/var/lib/vz/images/117/vm-117-disk-1.raw

Then when the VM was instructed to start it did not enable and use the image from my Qnap but instead used the image found under /var/lib/vz/images/117/vm-117-disk-1.raw despite the fact that the config pointed to the VM to use the image on my Qnap?????

Does this not qualify for a bug report?

 

[mir]

Just to proofe:
Jan 01 01:14:04 starting migration of VM 117 to node 'esx2' (192.168.2.9)
Jan 01 01:14:04 copying disk images
Jan 01 01:14:04 starting VM 117 on remote node 'esx2'
Jan 01 01:14:07 starting migration tunnel
Jan 01 01:14:07 starting online/live migration on port 60000
Jan 01 01:14:09 migration speed: 256.00 MB/s
Jan 01 01:14:09 migration status: completed
Jan 01 01:14:12 migration finished successfuly (duration 00:00:08)
TASK OK

And the config is still the same:
cat /etc/pve/qemu-server/117.conf
#eth0%3A 172.16.1.7
bootdisk: virtio0
cores: 1
cpu: qemu32
memory: 512
name: proxy
net0: virtio=1A:1D:E8:65:38:8C,bridge=vmbr1
ostype: l26
sockets: 1
virtio0: qnap_lvm:vm-117-disk-1,cache=writeback,size=4G

 

[dietmar]

Try

# qm rescan --vmid 117

After that you should see the unused disk in the config.

 

[mir]

I did that also but the disk did not show up. I would have been surprised if it in fact had turned up because this disk has never been assigned to this VM so how promox ever decided to use this disk as boot device beats me?

 

[dietmar]

I did that also but the disk did not show up.

Really? This is a bug then.

I would have been surprised if it in fact had turned up because this disk has never been assigned to this VM so how promox ever decided to use this disk as boot device beats me?

The disk name indicates that the disk belongs to VM 117!

 

[mir]

Really? This is a bug then.

Yes

The disk name indicates that the disk belongs to VM 117!

I know but it was something related to the test of my storage migration module. The disk has never been assigned to the VM nor has it ever been used for that VM.

 

[dietmar]

I know but it was something related to the test of my storage migration module. The disk has never been assigned to the VM nor has it ever been used for that VM.

The does not matter - the disk name is used to assign a disk to a VM.