Xen Project Q&A Forum: First Line Help for Simple Questions
This is your chance to ask questions and provide answers about basic use of the Xen Project software. For debugging problems and for more complex issues, consider using the xen-users mailing list instead. You can find information about xen-users under "HELP | Mailing Lists" in the navigation bar above.
I have searched information on it since a couple of days, but I am still stuck with it.
This is more a core linux question, but the issue also involves xen, has it is xen that defines the bridged interfaces. Or maybe this is a network misunderstanding on my side .
I am using xen 4.1.4 of the Debian Wheezy distribution, with network in bridge mode.
When a multicast message (here it was Upnp) is sent from the physical network, it can be seen in the eth1 and xenbr1 interfaces, but not in the interfaces inside the VMs. However tcpdump shows a number of 'dropped by interface' packets that is exactly equal to the number of expected packets.
I of course checked the iptables config.
Here follows some details:
Check on the xenbr1 (or eth1) interface (I am triggering an Upnp broadcast from an external physical device):
~# tcpdump -ni xenbr1 net 18.104.22.168/8 tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on xenbr1, link-type EN10MB (Ethernet), capture size 65535 bytes 15:35:16.545747 IP 192.168.0.70.38911 > 22.214.171.124.1900: UDP, length 98 15:35:16.643771 IP 192.168.0.70.38911 > 126.96.36.199.1900: UDP, length 98 2 packets captured 2 packets received by filter 0 packets dropped by kernel 2 packets dropped by interface
Check on the VM guest with the same test:
~# tcpdump -ni eth1 net 188.8.131.52/8 tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on eth1, link-type EN10MB (Ethernet), capture size 65535 bytes 0 packets captured 0 packets received by filter 0 packets dropped by kernel 2 packets dropped by interface
So I have insufficient knowledge to understand how to identify why I have dropped by interface packets, and what they are.
I also don't know how to tell the bridge to propagate those multicast packets to all VMs. This is what a bridged network is supposed to do...
I attached some outputs of my config. involved interfaces are eth1, vif1.1 and xenbr1.
You will see a lot of errors and drops in eth0, but this is another story... (I don't know why either, if you have a clue...)
(the .cfg of the vm, the ifconfig, iptables and sysctl params of the physical machine)
Thanks a lot in advance.
Did someone already succeed managing multicast messages inside xen VMs?
NB: I tested the same situation into a VirtualBox VM, in bridged network, and it worked just fine.
Accepted AnswerRussell PavlicekOffline0My apologies for not responding to this last summer when it was first posted.
An issue of this complexity is often best brought to the xen-users mailing list. There are a lot more people there who can help with detailed diagnosis.
This Q&A board is useful for general queries, but the more detailed matters are best brought to the mailing lists where many more of the developers regularly hang out.
Accepted Answer0Thanks, I will submit that on the mailing list as suggested. I chose the 'xen-users' list, is it the most appropriate?
I have built a newer machine with 3 network interfaces, updated version and tested it again (the test was slightly different as the fist broadcast packet directly arrives in the interface - in previous test it came from a wifi router); it acts the same...
I hope I will understand what is wrong with help from the mailing list guys, and also hope this won't bother too much.