I have the same result on vmware running a VM8 machine with vmxnet3 nic's.
Although something is mentioned in the debian-unstable merge ,
quote "+ Do not use start-stop-daemon and </dev/null to avoid blocking boot."
Implications go beyond booting. This should probably not have been implemented until this issue was fixed or a reasonable _temporary_ workaround found.
For example a PID for each running openvpn instance could be gotten , then one could loop netstat until no more ports linked to those pid's is open.
In extremis even a few seconds sleep before "stop" returns would probably work.
An actual fix would probably be best , but imo stop gap measures should be implemented before releasing if releasing is primordial.
I have the same result on vmware running a VM8 machine with vmxnet3 nic's.
Although something is mentioned in the debian-unstable merge ,
quote "+ Do not use start-stop-daemon and </dev/null to avoid blocking boot."
Implications go beyond booting. This should probably not have been implemented until this issue was fixed or a reasonable _temporary_ workaround found.
For example a PID for each running openvpn instance could be gotten , then one could loop netstat until no more ports linked to those pid's is open.
In extremis even a few seconds sleep before "stop" returns would probably work.
An actual fix would probably be best , but imo stop gap measures should be implemented before releasing if releasing is primordial.
With this:
Also affected.