As part of the last question I asked about firewall rules, I am looking at our firewall right now and monitoring the traffic. I'm looking at the traffic between VPN connections from our stores to a main server. These stores are all using the same application to communicate with the server. However, I'm looking at the server and it is receiving connections from our various stores, but every single store is communicating via a different port. So one store will be coming through port 4274. The other one will send it via port 4288. My point is, are applications specifically written in this way to prevent security breaches from happening by constantly randomising their port sequences so that they can't be 'guessed' by a malicious attacker?
And if that is the case, surely going back to the answers being given previously, this does warrant the ability for the 'ANY' ports to be open from site A to site B via VPN.