At the moment, the --root-dir argument (poorly named in my opinion) of the binary is not exposed on the node join command. We can add that (though I would prefer to call it something like --data-dir-path). We aren’t exposing --log-dir either on node join. Perhaps on the node join command, we can have:
--data-dir-path for the path of the data directory
--bin-dir-path for the path of the directory where the binary is (at the moment this is called --node-dir-path, but I think the word “node” is redundant in this context, since if we’re on a node join command, we know the argument relates to the node)
--log-dir-path for the path of the logs directory
That’s probably what I’d prefer. Not sure it will get done on this side of the year though. Maybe this week if I get time, but I’m gonna take a holiday after this week.
Well, I think setting up that arrangement of VMs is beyond the scope of what we could help you with here, but if you were interested in running a network where some nodes were running on machines on a cloud provider, we have a tool we use to do that:
This is what Josh uses to run his testnets. Basically what this tool does is, it spins up some machines on Digital Ocean and installs some nodes on them. Realistically, you’ll need a bit of technical competency to use it, but myself and probably others in here could help with some guidance around it.