[Xen-users] Problems using NAT in Xen 3.0

[Roberto Virga <rvirga@xxxxxxxxxxxxxxxxxxxxx>]

I'm trying to set up Xen 3.0 unstable on my machine(Athlon 64 X2 4200+,
running FC4/x86_64), and almost everything works well except domU
networking. Since connection to the internet is through an ADSL modem
(ppp0), I decided to use NAT rather than bridging. I used the network-nat
and vif-nat scripts from the tools/examples, changing 'eth0' with 'ppp0'.
Networking works somewhat, but with huge packet loss depending on the
protocols. Specifically, ping and ftp work without any problem. ssh
connections and dns lookups take a very long time to respond (~1 min.), but
eventually get a response, so not all the packets get lost. This is
particularly problematic with DNS, since lookups time out before they get a
response from the servers, and the way I found out they get a response at
all is to manually run 'telnet <DNS_server_IP> 53'.
This happens regardless if I try to connect to the internet or to just dom0,
which makes me believe it's not a problem with the IP tables configuration
(I've tried setting up a DNS server on dom0, with the same effect).
I've noticed the following odd thing running 'ip addr':

> 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>     inet 127.0.0.1/8 scope host lo
> 2: vif0.0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop
>     link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
> 3: veth0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop
>     link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
> 4: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
>     link/ether 00:13:d4:33:8c:71 brd ff:ff:ff:ff:ff:ff
> 6: vif1.0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue
>     link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
>     inet 192.168.1.1/24 brd 192.168.1.255 scope global vif1.0
> 7: ppp0: <POINTOPOINT,MULTICAST,NOARP,UP> mtu 1492 qdisc pfifo_fast qlen 3
>     link/ppp
>     inet 82.52.83.171 peer 192.168.100.1/32 scope global ppp0

In the above, vif0.0 (what's this for, anyway?) and vif1.0 have the same MAC
address (fe:ff:ff:ff:ff:ff), which it cannot be good. I've tried setting a
different MAC using the 'be_mac' option in the domU configuration file:

> vif = [ 'mac=aa:00:00:00:00:11, be_mac=aa:00:00:00:00:22,
ip=192.168.1.1/24' ]

but the option is ignored and the MAC is always set to fe:ff:ff:ff:ff:ff
regardless.

So my questions:
1. Has anybody managed to use Xen 3.0 unstable with NAT?
2. How do I set the MAC of the vif<n>.0 interfaces?
3. Do any of you have any suggestions of things I could try in order to
   diagnose/fix this problem? (I'm a newbie at Linux networking :-))

Thanks in advance for your help,

- Roberto



_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users