I have a quick example where this is not working correctly.

To start I have disabled the network-bridges script from /etc/xen/xend-config.sxp file. I followed the instructions in another forum. I basically replaced the line (network-script network-bridges) with (network-script network-dummy). This allows me to create my own bridge below.

I have two devices eth0 and eth1. I assign 0.0.0.0 to the two devices and then bridge them to joesb. I assign a public ip address to joesb and then bring up eth0 and eth1. I am then able to ping a server (192.168.0.1) that has ovm installed and I get a response back. Below is the following commands I use to accomplish this.

brctl addbr joesb
ifconfig eth0 0.0.0.0
ifconfig eth1 0.0.0.0
brctl addif joesb eth0
brctl addif joesb eth1
ifconfig joesb 192.168.0.100 netmask 255.255.255.0 up
ifconfig eth0 up
ifconfig eth1 up
ping 192.168.0.1

I then edit the /opt/ovs-agent*/OVSXXenStore.py the way you have specified it above. After rebooting or stopping the ovs-agent and restarting it I am still unable to see the bridge joesb in the drop down list.

A few questions:
Where and when do the functions xen_get_network_bridges and xen_get_network_bridges2 get called?
Why do you not use the system command brctl show to identify what bridges have been created on the system?
Is it necessary to have a peth device because it seems there are numerous if statements in the python code that search for a device that starts with the letter p?

I have heard that the python code in the agent will change to allow system administrators to create their own network scripts and devices. This will allow us to name our bridges anything we want. Is this code complete yet for the new release? If so can we pilot it?

 out of  found this valuable. Do you find this valuable?