Hi Alessandro,
maybe it might help to take a closer look at each machines mac addresses.
does this only happen on machines where vms are bound to more then one
bridge interface?
i think the kernel logs this when "nicely" dropping packets that appear
to come from "itself". so check mac addresses and try to specify
different mac address when setting up your bridges.
try to tcpdump network traffic between the affected nodes.
Am 18.11.2010 13:34, schrieb Alessandro Bono:
Hi
I have a problem stacking bonding + vlan + bridge, typically my servers
has two nic attached to two switch, one per nic, so I use bond in active-
passive configuration, on top of this I create vlans and a bridge for vlan
On ubuntu my network configuration are like this
auto bond0
iface bond0 inet manual
slaves eth0 eth1
bond_mode active-backup
bond_primary eth0
bond_miimon 100
auto br18
iface br18 inet static
address x.x.x.243
netmask 255.255.255.0
gateway x.x.x.254
bridge_ports bond0.18
bridge_fd 0
bridge_maxwait 0
auto br23
iface br23 inet manual
bridge_ports bond0.23
bridge_fd 0
bridge_maxwait 0
and so on
This machines are used as host for virtual machine, so vm are attached to
one or more of these bridge
If either eth0 and eth1 are up there are lots of this message in syslog
bond0.18: received packet with own address as source address
and only solution is to do
ip link set eth1 down
after this everything works without problem
note that this problem appear in random way and "seems" related to
presence of a vm that do bridge with his eth (normally for openvpn server)
I have this problem either in xen and kvm intallation, 2.6.24 and 2.6.32
respectively.
Googling I found lots of person with the same problem but without a clear
solution, any hint how to solve/debug?
thanks
--
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
