What happened?

Constellation server hosted on Server A. I want to tunnel Constellation to Server B and Server C.

Server A is on a separate network with it's own IP issued by the provider. Server B and C are on the same network, behind the same ISP provided address. I started by adding a device on Server A to connect to Server B, and had no issues. It was able to connect and showed the handshake messages received, the log also shows the tunnel was opened. I can ping server B and I get a response, server B can ping server A and get a response.

Next I added Server C to constellation and attempted to connect. The handshake message is sent by Server A to Server C according to Server A's logs. Server C attempts to send a handshake to Server A and it times out. Then the whole cycle repeats. I am not able to ping either server from either direction. The logs show that it's using the external IP and port 4242 to try and connect to both Server B and Server C.

What should have happened?

Server A should be able to connect to Server C, send data, and mesh it into the VPN.

How to reproduce the bug?

Adding 2 different Cosmos servers behind the same external IP address to Constellation reproduces the bug. I have reset the network, restarted each cosmos server, and readded them. The same issue was present.

Relevant log output

No response

Other details

I've followed existing documentation for setup, and have the appropriate ports open.

System details

Debian 12/Ubuntu 22.04
Virtual Machines
Cosmos .16
Constellation host in VPS