I can only set up the bridge interface from the command line using the bridge-start script. I can then start the service and my VPN works.
However, if I attempt to either modify my interface config files or run bridge-start with the OVPN service at boot, then my node becomes unreachable.
I've tried on CentOS 7.2 and Ubuntu 15, and the result is the same.
Currently I'm using Ubuntu, and setting up the interfaces like suggested here:
https://help.ubuntu.com/community/OpenVPN
I can edit the files and /etc/init.d/networking restart and have everything working fine, but as soon as I reboot the VM is lost (Connection Refused).
I suspect the problem is somehow related to the GCE DNS server on 169.254.169.254 becoming unavailable when I set up the bridge interface, but I can't be sure.
Probably not coincidentally, using the br0 interface causes the web based ssh client to cease functioning. Putty continues to work.
I did have the service working on a CentOS node as recently as yesterday. Something may be happening or changing upstream of me.
Last few lines of serial console output after rebooting to a locked out system are:
Mar 25 06:44:51 ns1 cloud-init[1086]: Cloud-init v. 0.7.7 finished at Fri, 25 Mar 2016 06:44:51 +0000. Datasource DataSourceNone. Up 14.15 seconds
Mar 25 06:44:51 ns1 cloud-init[1086]: 2016-03-25 06:44:51,524 - cc_final_message.py[WARNING]: Used fallback datasource
Mar 25 06:44:51 ns1 systemd[1]: Started Execute cloud user/final scripts.
Mar 25 06:44:53 ns1 kernel: [ 15.680040] br0: port 1(eth0) entered forwarding state
Ubuntu 15.10 test-server ttyS0
ns1 login: Mar 25 06:44:56 ns1 kernel: [ 18.816034] br0: port 2(tap0) entered forwarding state
I realize this is close to being outside the scope of the forums, but I thought there might be a chance someone here has run into this before.
If nothing else I may be able to file a bug report.
Thanks in advance.
