Description of problem: When switching between networks (eth0 -> wlan0, wlan0 -> eth0), virbr0 loses it's routing to the outside 'net. Version-Release number of selected component (if applicable): libvirt-0.4.4-2.fc9.x86_64 How reproducible: Always Steps to Reproduce: 1. Establish a link using NetworkManager. 2. Start a VM (I'm using KVM + Qemu). 3. From the VM ping an IP or hostname on the outside net (ping www.redhat.com). 4. Restart the NetworkManager connection (left-click on the NM applet and click on the interface that is currently running to re-negotiate with the network). 5. From the VM ping an IP or hostname on the outside net (ping www.redhat.com). 6. Fail! Actual results: Fail. Expected results: ping successful libvirtd should notice when the machine changes networks and recreate an appropriate routing table. Additional info:
Can't reproduce here Could you try running these commands before and after you restart the connection: $> brctl show $> iptables -L -v -n $> ps -ef | grep dnsmasq $> ifconfig -a $> cat /proc/sys/net/ipv4/ip_forward I suspect it might be related to bug #467687
Created attachment 330653 [details] before reconnecting w/ NM
Created attachment 330654 [details] after reconnecting w/ NM
I recently upgraded to F10 and can't reproduce the problem, either. *shrug* Take a look at the attached files - I don't see a smoking gun, but it sure sounds like bug #467687 and this are related. Cheers, Dan
(In reply to comment #4) > I recently upgraded to F10 and can't reproduce the problem, either. *shrug* > > Take a look at the attached files Ah, okay - these log files are from a system that the problem isn't occurring on > - I don't see a smoking gun, but it sure sounds like bug #467687 and this are > related. Yeah, I think we can be fairly certain that it was the same issue. Marking as duplicate, but do re-open if it comes back and looks different. *** This bug has been marked as a duplicate of bug 467687 ***