Feedback on security topic - IPSEC to be exact
A potential IPSEC partner said the following…
The destination server need not be on a private IP. VPN can be terminated on a public IP. The supposed VPN connection is depicted below:
XXX'shost–-XXX's Firewall--XXX's VPN Gateway---Public Internet----ABC's VPN Gateway---ABC host
VPN connections do not terminate on host machines; VPN connections terminate on the VPN gateways. Therefore, the VPN is required between XXX's VPN gateway and ABC's VPN Gateway. Between the host and the VPN gateway, the traffic is not encrypted. So, whether private or public is not the question, the question is what the objective of using public or private IP is on the user network. For XXX,the reason why we choose to have public IPs for host machines are:
1. We do not want to have Private IP address conflict that may arise now or in the future. OKAY, makes sense
2. We do not allow private IPs have access on the outside interface of our firewall. This is to prevent any possibility of spoofing attacks using private IPs. All access on this interface must be public. But isn't this why we are using IPSEC?!
3.Why we do not terminate VPN connections on our firewall is because we want our firewall to be able to filter all traffic using some of the basic intrusion prevention features it has. This will only be possible when traffic pass through or terminate on the firewall unencrypted. Maybe they need a second layer instead?
Based on those statements above, can pfSense 1.2 handle this? On the IPSEC connection, the end point is a public routable ip, and the actual host we will be connecting to will also have a PUBLIC IP?
Any other comment?