Apologies in advance that this is a bit off topic, but I'm desperate and did
see an xCursion related query the other day, and can't find anything
very useful anywhere else...
I am trying to arrange external access to an X based application running
on our DS20 Tru64 server - internally we run xCursion on our workstations
and
start (rexec) an xterm on the server , and then invoke the X based
application.
Assuming I can get the initial rexec of the xterm through the firewall,
I then have to get the X traffic itself through. My network manager wants to
know how the X traffic can be identified.
As far as I can see from observing internal sessions with netstat, a typical
session involves TCP port 6000 on the workstation - i.e. the X Server ,
and an arbitrary TCP port on the DS20 - i.e. the X client end.
(I assume that what this actually represents is a socket, since I know that
setting up a socket connection with some server process will result in
a pair of ports like that - on the server , whatever port the service was
listening on ,
plus an arbitrary port at the client end.)
However it doesn't look like recognising X traffic by TCP port usage will
work :
1) My network manager says he needs definite fixed port numbers, its
insufficient to
specify "port 6000 from machine X to port ???? on machine Y" - he needs
to
know the ????
2) There's every chance that the other end's firewall will be spoofing the
originating address and port of the workstation/X Server, to a different
address and port, so
even the port 6000 of the X Server may be mangled.
Anybody know how to get firewalls to recognise and allow X traffic ?
Another possibility I have read about is to run X through SSH - perhaps that
might
be the end result of this.
Thanks
Alan McCulloch
Bioinformatics Software Engineer
AgResearch NZ
Received on Mon Dec 18 2000 - 22:48:31 NZDT