openstack team mailing list archive

[Daniel Ellison <daniel@xxxxxxxxxx>]

Hi all,

I have Nova all set up on a single server and am able to start/stop/delete VM instances no problem. I have a bridge at br100 which sits on eth1 and is not connected to anything. eth0 is connected to the Internet. Before installing Openstack I was using KVM and virsh to manage my VMs. In order to do the Openstack install with fewer working parts, I brought down all KVM instances and deleted the br0 bridge they were using.

Everything works beautifully with respect to nova-network. Since I can't easily port my KVM instances to Openstack, I wanted to start them up again under virsh. I recreated the br0 bridge as it was before. So far so good. I can start my "legacy" VMs and all works as expected. There's only one issue, and I don't even know if it's important.

Before starting a Nova VM eth0 has no IP, which is expected as it's being covered by br0. But when I start one of the Nova VMs that has a floating IP, eth0 gains its IP! Everything seems to continue working, but it doesn't make sense to me.

I don't know if this is expected behaviour or I simply have things configured wrong. Since it all still works I'm not overly concerned, but it does bug me. If anyone has insight into this I would be grateful.

Thanks,
Daniel