Hi Toghether
i've got a bit a complicated setup. one of our partners like to have our traffic NATed and VPNed though one of their Address.
setup is as following:
[FONT="Courier New"]
Our Phys IP: 212.249.30.114
their phys IP: 212.249.30.113
their gateway: 212.249.30.113
their VPN GW: 212.190.80.140
their server: 212.190.80.137
our server: 212.55.190.140 (fw interface is 212.55.190.129)
our 'bogus' IP assigned from our partner: 195.129.120.33
[/FONT]
since this is a dedicated route to them (we have another ISP with default route) i just added static routes to the whole 212.190.80.0/24 setup. Physical Setup and logical setup works fine. i can reach their VPN GW and server as expected.
but they want to have our traffic encrypted since we're using a plaintext application. and in order to route our traffic back to us they want us to send them our traffic NATed through our bogus IP
the VPN setup looks like this and comes up as expected:
[FONT="Courier New"]
195.129.120.33--212.249.30.114212.190.80.140---212.190.80.137
(bogus IP)-----(our Iface)(their VPN)------(des.srv)[/FONT]
now so that our server can reach their server through the VPN i added a NAT rule which say's:
[FONT="Courier New"]From: 212.55.190.140 Service ALL
Dest.: 212.190.80.137
SNAT: 195.129.120.33
DNAT: None
Service Tranlation: None[/FONT]
which should translate the SourceIP 212.55.190.140 to our bogusip, 195.129.120.33.
this works fine. but when i do a traceroute from our server is see too many hops between.
what i expected to see:
[FONT="Courier New"]1 - 212.55.190.129 (our Firewall)
2 - 212.190.80.137 (their server)[/FONT]
what i see:
[FONT="Courier New"]1 - 212.55.190.129 (our Firewall)
2 - 212.249.30.113 (their gateway)
3 - some router
4 - some router
5 - some router
6 - some router
7 - timeout
8 - timeout
9 - timeout
...[/FONT]
so. i expected to have max 3 hosts since we're going through the tunnel and the TTL shouldn be decreased since my packets arent altered between the two VPN endpoints.
but what i actually see is the complete route with every router between our and their VPN endpoints.
so my concrete question:
could it be that the VPN rule takes place before the NAT rule?
The VPN sees that the packet doesn't match the parameters for routing this packet though the VPN and give's it further. then the NAT rule is applied and the packet is sent as usual plain out to the interface.
when yes, how can i force that our server's packets are NATed BEFORE the VPN, an then send through the VPN?
changing our server's IP address is not an option since we running there a complex setup as well.
any help would be appreciated
cheers
josh
This thread was automatically locked due to age.