I am having an issue with my setup where the mention service will wait and eventually hang on bootup. I think it is because I have a 10G nic with 2 hosts that are direct attached (static IPs, no gateway, no DNS), but not always powered on. What I believe is happening is the service is waiting for the links to come up, but they don't since the endpoints are down. This causes the service to eventually fail.
An interesting observation is that the primary interface (a 1g connection that has a static IP in 192.168.150/24) comes up right away, is pingable, I can ssh into it, and I can bring up the web interface here. Other services remain down, however, while wait online is waiting including all the docker networking, smb, and ftp. As soon as the service fails, everything else comes up and the server is fine with the exception of me getting emails saying rrcached and collectd taking too long to start and restarting.
Removing the direct attached 10G interfaces from OMV (but leaving the nic in and everything) solves this issue and the system comes up fine and the service doesn't hang and fail. Also, this never happened in OMV4, but I guess the new systemd networking in OMV5 is causing this.
After reading about a bit, I found an option I can add to the systemd-networkd-wait-online.service unit telling it to ignore certain interfaces, or to just pass when any interface is up. My question is what is the best way of doing this in OMV 5? Should I edit the unit file in /etc/systemd/system/network-online.target.wants/systemd-networkd-wait-online.service directly, or is there an environment variable I should add to /etc/default/openmediavault and re-deploy the salt env?
If anyone that can verify they are seeing the same thing with interfaces that are waiting for a link, but don't have one on boot it would be great.