Hello,
I have IP masquerading working fine on my Redhat 5.0 system, for all
normal applications. My setup is a local LAN on the 192.168.99.n net,
including my firewall which is connected to the internet via cable-modem
(which is connected via a second ethernet card in the firewall machine).
I have one application (a game: Battlezone) that I run on a machine
inside the firewall. It opens a single UDP port and talks to a list of
game servers. From this list I choose a server, and log on to it. This
is all still using the same port on the private machine. Once logged on
to a server, my local machine 'pings' (via UDP) all other players on
that server, and they ping me. My problem is that the firewall assigns
a new port to every players' IP address as I ping them, while the
players' machines are trying to ping me on the port that the server
knows about. Whenever any player tries to ping me on the server's port,
the firewall denies it.
I've done some reverse engineering based on captured traffic from when
thet game machine was directly connected to the internet. I've also
fiddled around with writing a proxy, based on the ip_masq_quake code.
Is it possible to configure ipfwadm or have a proxy that will masquerade
all traffic using a SINGLE port on the firewall, regardless of the
destination IP/port? And is it then possible to forward the packets
sent to this masquerade port to the private machine's port, also
regardless of the source of the packet?
If such a scheme is possible, I think I'll need a proxy anyway, as the
game's protocol embeds IP addresses and port numbers in the packet's
payload, but my experimental proxy can handle this (admittedly in a
crude way).
Any help at all would be appreciated,
Jon Wurtz
[EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
For daily digest info, email [EMAIL PROTECTED]
