Take the 2-minute tour ×
Ask Ubuntu is a question and answer site for Ubuntu users and developers. It's 100% free, no registration required.

I can't for the life of me get bonding and bridging to work for the KVM setup I'm building. I'm using a fresh install (not an upgrade) of Ubuntu Server 10.10. I have 4 NICs on the same subnet (two intended for each of my two VMs). I'm trying to achieve the setup that Uthark describes here.

But following his guidelines didn't work for me. My eth0 and eth1 did not come up, and "brctl show" showed that br0 didn't have any interfaces (the bond). I assumed it didn't work because he's using 10.4, and this article says there's a recent change in bonding:

[I can't post more than one hyperlink per post because I'm a newbie.]

I had to use this article to get my interfaces to work at all on the same subnet, which is why I have the post-up lines on some of my interfaces:

[I can't post more than one hyperlink per post because I'm a newbie.]

I installed ifenslave and ethtool. I also created /etc/modprobe.d/aliases.conf with the following content:

alias bond0 bonding
options bonding mode=6 miimon=100 downdelay=200 updelay=200

And I included "bonding" in /etc/modules

So, after several approaches, here is my latest interfaces file:

auto lo
iface lo inet loopback

auto eth5
iface eth5 inet manual

auto br5
iface br5 inet static
post-up /sbin/ip rule add from [network].79 lookup 10
post-up /sbin/ip route add table 10 default via [network].1 src [network].79 dev br5
address [network].79
netmask 255.255.255.0
network [network].0
broadcast [network].255
gateway [network].1
bridge_ports eth5
bridge_stp off
bridge_fd 0
bridge_maxwait 0

auto eth2
iface eth2 inet manual

auto br2
iface br2 inet static
post-up /sbin/ip rule add from [network].78 lookup 11
post-up /sbin/ip route add table 11 default via [network].1 src [network].78 dev br2
address [network].78
netmask 255.255.255.0
network [network].0
broadcast [network].255
gateway [network].1
bridge_ports eth2
bridge_stp off
bridge_fd 0
bridge_maxwait 0

iface eth0 inet manual

iface eth1 inet manual

auto bond0
iface bond0 inet static
bond_miimon 100
bond_mode balance-alb
up /sbin/ifenslave bond0 eth0 eth1
down /sbin/ifenslave -d bond0 eth0 eth1

auto br0
iface br0 inet static
address [network].60
netmask 255.255.255.0
network [network].0
broadcast [network].255
gateway [network].1
bridge_ports bond0

eth2, eth5, br2, and br5 all seem to be working fine.

The only other thing I could find that looked suspicious is an error regarding bonding in /var/log/messages:

kernel: [ 3.828684] bonding: Warning: either miimon or arp_interval and arp_ip_target module parameters must be specified, otherwise bonding will not detect link failures! see bonding.txt for details.

even though there is a bond-miimon line in /etc/network/interfaces (if that's what they're talking about).

Also, the bond seems to go in and out of promiscuous mode several times on boot:

Jan 20 14:19:02 kvmhost kernel: [ 3.902378] device bond0 entered promiscuous mode
Jan 20 14:19:02 kvmhost kernel: [ 3.902390] device bond0 left promiscuous mode
Jan 20 14:19:02 kvmhost kernel: [ 3.902393] device bond0 entered promiscuous mode
Jan 20 14:19:02 kvmhost kernel: [ 3.902397] device bond0 left promiscuous mode
Jan 20 14:19:03 kvmhost kernel: [ 4.998990] device bond0 entered promiscuous mode
Jan 20 14:19:03 kvmhost kernel: [ 4.999005] device bond0 left promiscuous mode
Jan 20 14:19:03 kvmhost kernel: [ 4.999008] device bond0 entered promiscuous mode
Jan 20 14:19:03 kvmhost kernel: [ 4.999012] device bond0 left promiscuous mode

Any advice would be greatly appreciated. It seems that this must be possible, based on other posts, but I can't see what I'm doing wrong.

Thanks.

share|improve this question

closed as too localized by Marco Ceppi Jan 13 '12 at 13:18

This question is unlikely to help any future visitors; it is only relevant to a small geographic area, a specific moment in time, or an extraordinarily narrow situation that is not generally applicable to the worldwide audience of the internet. For help making this question more broadly applicable, visit the help center.If this question can be reworded to fit the rules in the help center, please edit the question.

    
If you add the URLs in comments here I'll put them in your question until you have enough rep to leave links. –  Jorge Castro Feb 1 '11 at 0:11
add comment

1 Answer 1

I ran into a similar problem a while back with Debian (4.something I think?) and it took me forever to realise the cisco catalyst switch's ports the NICs were plugged into didn't support the right configuration to allow this type of bonding. The bond0 would continuously cycle same as shown in your snippet. The version of IOS needed wasn't supported on that model Catalyst. Just something to check...

share|improve this answer
add comment

Not the answer you're looking for? Browse other questions tagged or ask your own question.