There is no difference in terms of performance. Having many peers is cheap; it’s total traffic that matters.
This one is trickier. If the clients are alike, then usually a single interface is used. When there are special considerations like the clients having addresses allocated from different subnets, require different routing considerations, are members of different VRFs, etc. creating more than one interface may be appropriate.
EDIT: In terms of no. 3, your diagram shows the simple one interface case.
Security wise multiple interfaces might be better but that would be more from an efficiency point of view. In firewall you can do what you want with interface or peer. Same with routing. Usually alike connections go on the same itf.
Performance wise it doesn’t matter that much.
As long as your central hub has enough processing power to handle the load, you should be fine.
I have a couple of setups carrying multiple active peers ( mostly management control and EOIP tunnels to allow ROMON). Even multiple interfaces.
Each time a central server allowing connections on a single UDP port per interface.
I created two tunnels. The first one is working normally with internet access. The second one only shows received packets, but no packets are being transmitted.
here is my config for both tunnels
It goes without saying that you shouldn’t share keys, especially the private sides. Regenerate them once your config is working correctly.
I have no idea what this means:
Endpoint = x.x.x.x:50.x.x.x
Other then that, the fact that you are receiving packets correctly usually means that the Mikrotik-side config is incorrect. Show it, without that I can only guess as to what exactly is wrong, but the usual one is misconfiguration of the allowed-ips field. For the first peer it should be 172.144.0.6/32, for the second 172.44.0.4/32
Hello @lurker888.
I removed them…Thank you.
So, I encountered the following problem:
First, I created a WireGuard tunnel with only one peer, and the connection worked well.
Then, when I added another peer to the same interface, the connection stopped working.
It seems that the connection only works for one peer at a time.
Is there anything I need to enable or configure to make the connection work for both peers?
here is the config for both peer and interface
the first peer
Look at it from another way…
If you put /24, you allow the complete subnet.
How does the server ( term used for ease of explanation) know then where to send a packet to ?
With /32 there is no confusion. It’s that address and nothing else.
On peer side, you can use /24 since everything from that subnet has to go into the tunnel.
So, on the MikroTik side, in IP → Addresses, I should assign the WireGuard interface an IP like 172.20.0.1/32, right?
Then for the peers (also on the MikroTik), I should set the allowed addresses as:
172.20.0.2/32 for the first peer
172.20.0.3/32 for the second peer
On the client side (outside the router), do I also set their IPs as /32 like 172.20.0.2/32 and 172.20.0.3/32, or can I use something like /24 instead?
I tried as you mention, still one of them work but the other doesn’t
Wireguard interface IP /24
both peer in the router /32
both peer outside the router /24
one of them work like normal bit the other not.
I will try on another newly created Wireguard interface and give you the output.
From terminal
/ export file=anynameyouwish
Move file to your pc
Obfuscate sensitive parts and post back between code quotes for easier readability.
Make sure it’s clear for us when you change things, something is there and use that same translation where it appears again.
You don’t have a working setup and you came here to ask for help.
Then why do you leave out parts which are important for us to see where the problem might be ?
Where is the interface definition for wireguard ?
As I asked, how are we supposed to know which keys are used on which peer if you simply blank them out ? How do we know what is what ?
And please post config between code quotes. I especially had to install a text editor on my tablet for opening that file.
If it was, as asked, between code quotes, I could immediatelly see it in the post.