Should the auto-healing retry indefinitely?

[smolyakoff]

I did something stupid and ran gluetun with network_mode: host on the remote Synology 🤦 As a result I've lost inbound connectivity both with Zerotier and Synology QuickConnect (which I used as backup method if I mess up with Zerotier). I see that some services on the NAS are still running and connect to remote peers, so outbound connectivity is somehow working there.

Is there any way the container would stop if VPN is never connected (it didn't have restart parameter specified)? Asking, because I think it would be generally useful failure mode if the container just exits if network is unavailable for some time with failing healthcheck.