RE: ICMP pokes holes in firewalls... (fwd from [email protected])

----- Forwarded message from Daniel Chemko <[email protected]> -----

From: "Daniel Chemko" <[email protected]>
Date: Thu, 25 Sep 2003 15:05:07 -0700
To: "H D Moore" <[email protected]>,
	<[email protected]>, <[email protected]>
Subject: RE: ICMP pokes holes in firewalls...

>NAT gateway has been 
>detected as a ignore-the-source UDP forwarder

2.4 kernels: NAT doesn't work without ip_conntrack, and ip_conntrack
always keeps track of source IP addresses (hence its function). I can't
think of a situation for any Linux machine which allows inbound UDP
replies from other sources. Spoofing the original sender's address is a
different story, but that is pandemic of any stateless AND insecure
protocol.

>I posted about this in March of 2000, the kernel development team
response 
>was that many RPC services require this functionality and it would not
be 
>fixed. The reason is that many UDP-based RPC services will respond back

>to requests from an alternative interface using a different IP address 
>entirely.

Just recently someone has written a conntrack handler to traverse
firewalls with RPC as you describe. No leaks to my knowledge, although I
am not too familiar with this module.

----- End forwarded message -----

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2-rc1-SuSE (GNU/Linux)

iD8DBQE/dBd5dbAkQ4sp9r4RAjWRAKC4SJTyrVBGCHpb+i4s4K/25yHcRACfWNke
CdiCGSZZcbtbCk8mY9sfDGY=
=hb2b
-----END PGP SIGNATURE-----