Just a heads up that I had the same problem -- thanks to @WarezMonkey for finding the root cause. Removing this line from /etc/daemon_mgr.conf seems to do the trick (different number in my daemon list apparently, but same idea):
DAEMON65 = openvpn, stop, /usr/sbin/openvpn
When I add the line back, things break after about 30 seconds, manifesting as a lot of entries like this in deluge:
Error: Cannot assign requested address
And on the docker logs:
[info] Deluge listening interface IP 10.8.10.6 and VPN provider IP 10.31.10.6 different, marking for reconfigure
I'm suspecting the reason it wasn't an issue for other users in the past may have to do with QNAP switching the default network mode on Docker to NAT instead of bridge at some point, but I'm not sure about that.
If it helps, I'm on a QNAP TS-251 running 4.3.6.0805, on ContainerStation 1.9.3590.
I also suspect if @WarezMonkey is right about how daemon_mgr works, simply renaming the docker image's openvpn binary to something else would do the trick, but that'd also involve updating all the scripts that call into it (and possibly os-level stuff in the image?)
Thanks for the hard work here; this beats the heck out of the hacked-together VM image I was using for this purpose before!
