[PVE-User] help please - upgrade v1.6 to v1.7 not so fine...

[Dietmar Maurer]

just a guess, seem the interface name gets too long:

we start with "ifname=vmtab5207007i0d0"

but kvm calls the script with "vmtab5207007i0d"

as workaround choose a smaller vm id -  does that help?

- Dietmar

From: pve-user-bounces at pve.proxmox.com [mailto:pve-user-bounces at pve.proxmox.com] On Behalf Of Harald Kapper
Sent: Freitag, 03. Dezember 2010 12:56
To: pve-user at pve.proxmox.com
Subject: [PVE-User] help please - upgrade v1.6 to v1.7 not so fine...

Hi
We got a currently working machine that goes in terms of vmbr like this:

bridge name     bridge id               STP enabled     interfaces
vmbr0           8000.00215e6725ce       no              eth0
                                                        vmtab5207006i0
vmbr1           8000.00215e6725cf       no              eth1.2305
                                                        vmtab5207006i1
vmbr2           8000.00215e6725cf       no              eth1.322
                                                        vmtab5207006i2

then there is a machine that has been upgraded to v1.7 fully that shows:
bridge name     bridge id               STP enabled     interfaces
vmbr0           8000.00215e6777b8       no              eth0
vmbr1           8000.00215e6777b9       no              eth1.2305
vmbr2           8000.00215e6777b9       no              eth1.322

if on this machine now a kvm-box tries to start it goes like this (and badly wrong) :
qm start 5207007
got strange interface name 'vmtab5207007i0d'
/var/lib/qemu-server/bridge-vlan: could not launch network script
kvm: -netdev type=tap,id=vlan0d0,ifname=vmtab5207007i0d0,script=/var/lib/qemu-server/bridge-vlan: Device 'tap' could not be initialized
start failed: command '/usr/bin/kvm -monitor unix:/var/run/qemu-server/5207007.mon,server,nowait -vnc unix:/var/run/qemu-server/5207007.vnc,password -pidfile /var/run/qemu-server/5207007.pid -daemonize -usbdevice tablet -name db-slave -smp sockets=1,cores=8 -nodefaults -boot menu=on,order=cd -vga cirrus -tdf -k de -drive if=ide,index=2,media=cdrom -drive file=/var/lib/vz/images/5207007/vm-5207007-disk-1.raw,if=virtio,index=0,boot=on -drive file=/var/lib/vz/images/5207007/vm-5207007-disk-2.raw,if=virtio,index=1 -drive file=/var/lib/vz/images/5207007/vm-5207007-disk-3.raw,if=virtio,index=2 -m 8192 -netdev type=tap,id=vlan0d0,ifname=vmtab5207007i0d0,script=/var/lib/qemu-server/bridge-vlan -device e1000,mac=FE:FF:0A:D2:02:0F,netdev=vlan0d0 -netdev type=tap,id=vlan1d0,ifname=vmtab5207007i1d0,script=/var/lib/qemu-server/bridge-vlan -device e1000,mac=02:FA:25:3A:71:7E,netdev=vlan1d0 -netdev type=tap,id=vlan2d0,ifname=vmtab5207007i2d0,script=/var/lib/qemu-server/bridge-vlan -device e1000,mac=2A:2C:99:17:83:07,netdev=vlan2d0' failed with exit code 1

we finally tried to reboot the whole box - nothing changed, it won't fix and start (we've seen before the need to restart bridges, but didn't help).

Any help appreciated as we have all boxes in this cluster configured like this I see hell breaking loose.

Regards
Harald Kapper / kapper.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.proxmox.com/pipermail/pve-user/attachments/20101203/5f95c66a/attachment.htm>