systemd target representing configured system

[tjkirch]

We need a systemd target representing the point at which all userdata (moondog) and dynamic (sundog) configuration is complete. This is what services like k8s and EKS will depend on, because they need the user's cluster settings to be fully applied to the system before starting.

• Create target
• Update k8s/EKS service units to depend on it, in addition to network or other existing dependencies

[zmrow]

My vote is to update moondog and sundog to not commit settings changes.

I don't like the implicit knowledge involved by just updating moondog. A simple atomic unit that does the commit is more in line with the rest of the single-use tools we are building.

[zmrow]

Also, I'm happy to update moondog and sundog and leave them in a PR to merge whenever we're ready.

[tjkirch]

We don't actually have to change the commit behavior in the same issue, because the services depending on sundog (with the new ordering from this issue) won't have started yet. So, it doesn't matter if we write config twice and "restart" the service, since systemctl won't "restart" a service that hasn't started yet, in the way that most sysv init scripts would.

I'm going to split that into another issue. We have to come up with a name for the thing that commits at boot, anyway, and naming is hard.

[tjkirch]

Fixed in #96