On Tue, Nov 24, 2020 at 9:41 PM Adam Mizerski
With wireguard, there's no server and client. Every node of this VPN is an equal peer. Just some of them must have static, well known IP address.
thanks for these basic infos and steps in this thread, also to the other fellow posters. as always, thanks to everybody participating here. i wasnt aware of such a rather-node than client-server concept. So in my case, for example the site-to-site thing i am trying to get into. would it suffice to have a linux box on both disjunct networks, they both reach each other via the internet public routed ip (dynamic dns provider and the like) and the lan side participants (windows machines) would only connect their wireguard to the locally available linux server, and these two linux servers via wireguard relaying all information among all windows stations in both lans, and thus forming a wireguard virtual lan where all participating stations would be members? so i dont have to wireguard-connect all stations with all other wireguard stations, right? in their respective configs. only peer the local lan1 stations with the linux wg machine in lan1, and the local lan2 stations with the local lan2 linux machine, and the rest flows naturally from linux-wg1 to linux-wg2 acting as routers in the wg layer?
I have some experience with that. Feel free to ask, if you have any trouble.
will see what i can make of it. actually maybe i would need to provide a route concept for non-wireguard capable machinery and devices that they all can be in this bridged-sort of overall-lan brought to life with this wg concept? would that be possible? then maybe i would only need those two linux1 linux2 machines with wg, and the rest targeting their packets to these two gateways locally or route only specific packets through those two devices? TY.