[Xen-users] eth0 disabled when xend starts up

["Davidson Rodrigues Paulo" <davidsonpaulo@xxxxxxxxx>]

Hi,

I was using xend 3.0.3 on CentOS 5 so I decided upgrade it to 3.1.2.
Before the update, I was using the following network configuration:

 (network-script 'network-bridge')
 (vif-script 'vif-bridge')

Where bridge's name was 'xenbr0' and netdev was 'eth0'.

After I upgraded xen, eth0 was disabled when xend was started. Running
'service network restart' enabled it up again. So, I changed my
xend-config.sxp as follows:

 (network-script 'network-bridge netdev=eth0'

So my pain started. When xend starts, eth0 was totally removed from
system, which can't detect it anymore, even stopping xend and running
'service network restart'. Moreover, xend created a tmpbridge which
assumed the IP address previosly used by eth0, but the network can't
be reached.

Analising /etc/xen/scripts/network-bridge I've found the correct way
to configure a bridge to use with Xen:

 # /etc/xen/scripts/network-bridge start netdev=eth0 bridge=eth0

Running that command creates a peth0 interface and a eth0 bridge, the
IP previously used by eth0 is preserved and the network can be
reached. Changing 'start' by 'stop' the network configuration is
restored for it's defaults, all fine. So I've changed xend-config.sxp:

 (network-script 'network-bridge netdev=eth0 bridge=eth0')
 (vif-script 'vif-bridge bridge=eth0')

So, I've started xend and for my surprise I got the same result, that
is, a xenbr0 and a tmpbridge bridges, a disabled eth0 and a network
out of reach.

It's the second time I face this type of problem. On the first time I
was using Fedora 8, I was with no patience so I have just reinstalled
the system, but I hated to do that, I want to know how to fix this
problem.

If you can help me in any form, I thank you.

Regards,

--
Davidson Paulo
http://fedoraproject.org/wiki/DavidsonPaulo

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