bridge / qemu

eantoranz · 12-31-2012, 02:39 PM

Hi!

I have this environment where I want to set up a pacemaker-based cluster.

Now, I'm noticing that either cluster node can't see the other node as active. After checking for a little bit what's going on I noticed that multicast traffic that is reaching each of the nodes of the cluster does not carry the src IP of the other node but the IP of the physical node where I'm running the qemu instances (physical host is running on kubuntu 12.10, qemu instances are running ubuntu server 12.04.1). Is there something I have to set up in the bridge or something so that they can see traffic as coming from the other node?

By the way, when nodes ping each other (or some other traffic) it does show with the right IP address. Why is multicast traffic different?

I won't swear on it but I have another set up working on a 12.04 host (running 10.04 servers) that is running fine with each node seeing the other without a hitch.

Code:

$ sudo brctl show
bridge name     bridge id               STP enabled     interfaces
puente          8000.9ee02eddbb94       no              tuncluster1
                                                        tuncluster2
                                                        tunsan
$ sudo ip link show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN mode DEFAULT 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state DOWN mode DEFAULT qlen 1000
    link/ether 00:21:9b:ee:ab:fe brd ff:ff:ff:ff:ff:ff
3: eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DORMANT qlen 1000
    link/ether 00:23:4e:6d:36:c1 brd ff:ff:ff:ff:ff:ff
4: puente: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT 
    link/ether 9e:e0:2e:dd:bb:94 brd ff:ff:ff:ff:ff:ff
5: tunsan: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master puente state DOWN mode DEFAULT qlen 500
    link/ether 9e:e0:2e:dd:bb:94 brd ff:ff:ff:ff:ff:ff
6: tuncluster1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master puente state UP mode DEFAULT qlen 500
    link/ether fe:31:82:2f:12:36 brd ff:ff:ff:ff:ff:ff
7: tuncluster2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master puente state UP mode DEFAULT qlen 500
    link/ether d6:dc:a3:3e:cf:3f brd ff:ff:ff:ff:ff:ff
$ sudo ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state DOWN qlen 1000
    link/ether 00:21:9b:ee:ab:fe brd ff:ff:ff:ff:ff:ff
3: eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:23:4e:6d:36:c1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.5/24 brd 192.168.1.255 scope global eth2
    inet6 fe80::223:4eff:fe6d:36c1/64 scope link 
       valid_lft forever preferred_lft forever
4: puente: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP 
    link/ether 9e:e0:2e:dd:bb:94 brd ff:ff:ff:ff:ff:ff
    inet 192.168.55.1/24 scope global puente
    inet6 fe80::281c:e6ff:fe7a:8e9a/64 scope link 
       valid_lft forever preferred_lft forever
5: tunsan: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master puente state DOWN qlen 500
    link/ether 9e:e0:2e:dd:bb:94 brd ff:ff:ff:ff:ff:ff
6: tuncluster1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master puente state UP qlen 500
    link/ether fe:31:82:2f:12:36 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc31:82ff:fe2f:1236/64 scope link 
       valid_lft forever preferred_lft forever
7: tuncluster2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master puente state UP qlen 500
    link/ether d6:dc:a3:3e:cf:3f brd ff:ff:ff:ff:ff:ff
    inet6 fe80::d4dc:a3ff:fe3e:cf3f/64 scope link 
       valid_lft forever preferred_lft forever

eantoranz · 12-31-2012, 06:46 PM

I bet it's something nagging around there. I shutdown everything and then when I restarted my computer and them the virtual machines, both computers were able to see each other. Then I shut down corosync on one of the nodes and then after a while I started it again and now I the nodes can¿t get any multicast traffic coming from the other node (not even with a natted source address).

What could be the problem here?