Hello,
I set up proxmox 1.9 on Debian lenny with DRBD.
Everything is working fine, except live migration with a virtio Ethernet device.
It works using an e1000 network card. And it even works if i configure a virtio Ethernet device but don't install the driver in Windows (my test machine is a Windows 2008R2 installation). I tried different Ethernet virtio drivers which are around the net. Especially driver ISO from Fedora (http://alt.fedoraproject.org/pub/alt/virtio-win/latest/images/) and the WHQL certified version from Redhat (tried verison 1.2.0 and 1.3.3).
I suspect this is a KVM bug, but I'm not sure so I ask here first. I get following output in syslog on unsuccessful migration:
Source node:
Oct 3 16:41:09 pve1 pvedaemon[7033]: migrating VM 101 from node 1 (localhost) to node 2 (192.168.7.202)
Oct 3 16:41:09 pve1 qmigrate[7034]: starting migration of VM 101 to host '192.168.7.202'
Oct 3 16:41:09 pve1 qmigrate[7034]: copying disk images
Oct 3 16:41:09 pve1 qmigrate[7034]: starting VM on remote host '192.168.7.202'
Oct 3 16:41:10 pve1 qmigrate[7034]: starting migration tunnel
Oct 3 16:41:10 pve1 qmigrate[7034]: starting online/live migration
Oct 3 16:41:10 pve1 qmigrate[7034]: VM 101 monitor command 'migrate -d "tcp:localhost:60000"'
Oct 3 16:41:10 pve1 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
Oct 3 16:41:10 pve1 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
Oct 3 16:41:12 pve1 qmigrate[7034]: VM 101 monitor command failed - VM not running
Oct 3 16:41:12 pve1 qmigrate[7034]: VM 101 already stopped
Oct 3 16:41:12 pve1 qmigrate[7034]: online migrate failure - VM not running
Oct 3 16:41:12 pve1 qmigrate[7034]: migration finished with problems (duration 00:00:03)
Oct 3 16:41:12 pve1 pvedaemon[7033]: VM 101 migration failed -
Target node:
Oct 3 16:41:10 pve2 qm[7707]: VM 101 start
Oct 3 16:41:10 pve2 kernel: device tap101i0d0 entered promiscuous mode
Oct 3 16:41:10 pve2 kernel: vmbr0: port 2(tap101i0d0) entering forwarding state
Oct 3 16:41:11 pve2 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
Oct 3 16:41:11 pve2 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
PVE Version is:
pve1:~# pveversion -v
pve-manager: 1.9-24 (pve-manager/1.9/6542)
running kernel: 2.6.32-6-pve
proxmox-ve-2.6.32: 1.9-47
pve-kernel-2.6.32-6-pve: 2.6.32-47
qemu-server: 1.1-32
pve-firmware: 1.0-14
libpve-storage-perl: 1.0-19
vncterm: 0.9-2
vzctl: 3.0.29-2pve1
vzdump: 1.2-16
vzprocps: 2.0.11-2
vzquota: 3.0.11-1
pve-qemu-kvm: 0.15.0-1
ksm-control-daemon: 1.0-6
pve1:~#
After failed migration the VM is registered on the target node but "stopped".
I hope anybody has a hint where to investigate further. I'd really like to use the virtio driver because networking performance is much better compared to e1000.
best regards,
Christoph
I set up proxmox 1.9 on Debian lenny with DRBD.
Everything is working fine, except live migration with a virtio Ethernet device.
It works using an e1000 network card. And it even works if i configure a virtio Ethernet device but don't install the driver in Windows (my test machine is a Windows 2008R2 installation). I tried different Ethernet virtio drivers which are around the net. Especially driver ISO from Fedora (http://alt.fedoraproject.org/pub/alt/virtio-win/latest/images/) and the WHQL certified version from Redhat (tried verison 1.2.0 and 1.3.3).
I suspect this is a KVM bug, but I'm not sure so I ask here first. I get following output in syslog on unsuccessful migration:
Source node:
Oct 3 16:41:09 pve1 pvedaemon[7033]: migrating VM 101 from node 1 (localhost) to node 2 (192.168.7.202)
Oct 3 16:41:09 pve1 qmigrate[7034]: starting migration of VM 101 to host '192.168.7.202'
Oct 3 16:41:09 pve1 qmigrate[7034]: copying disk images
Oct 3 16:41:09 pve1 qmigrate[7034]: starting VM on remote host '192.168.7.202'
Oct 3 16:41:10 pve1 qmigrate[7034]: starting migration tunnel
Oct 3 16:41:10 pve1 qmigrate[7034]: starting online/live migration
Oct 3 16:41:10 pve1 qmigrate[7034]: VM 101 monitor command 'migrate -d "tcp:localhost:60000"'
Oct 3 16:41:10 pve1 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
Oct 3 16:41:10 pve1 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
Oct 3 16:41:12 pve1 qmigrate[7034]: VM 101 monitor command failed - VM not running
Oct 3 16:41:12 pve1 qmigrate[7034]: VM 101 already stopped
Oct 3 16:41:12 pve1 qmigrate[7034]: online migrate failure - VM not running
Oct 3 16:41:12 pve1 qmigrate[7034]: migration finished with problems (duration 00:00:03)
Oct 3 16:41:12 pve1 pvedaemon[7033]: VM 101 migration failed -
Target node:
Oct 3 16:41:10 pve2 qm[7707]: VM 101 start
Oct 3 16:41:10 pve2 kernel: device tap101i0d0 entered promiscuous mode
Oct 3 16:41:10 pve2 kernel: vmbr0: port 2(tap101i0d0) entering forwarding state
Oct 3 16:41:11 pve2 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
Oct 3 16:41:11 pve2 kernel: vmbr0: port 2(tap101i0d0) entering disabled state
PVE Version is:
pve1:~# pveversion -v
pve-manager: 1.9-24 (pve-manager/1.9/6542)
running kernel: 2.6.32-6-pve
proxmox-ve-2.6.32: 1.9-47
pve-kernel-2.6.32-6-pve: 2.6.32-47
qemu-server: 1.1-32
pve-firmware: 1.0-14
libpve-storage-perl: 1.0-19
vncterm: 0.9-2
vzctl: 3.0.29-2pve1
vzdump: 1.2-16
vzprocps: 2.0.11-2
vzquota: 3.0.11-1
pve-qemu-kvm: 0.15.0-1
ksm-control-daemon: 1.0-6
pve1:~#
After failed migration the VM is registered on the target node but "stopped".
I hope anybody has a hint where to investigate further. I'd really like to use the virtio driver because networking performance is much better compared to e1000.
best regards,
Christoph