- From: Keith Moore <moore@cs.utk.edu>
- Date: Mon, 02 Dec 2002 14:44:34 -0500
- To: Patrik Fältström <paf@cisco.com>
- cc: Brian E Carpenter <brian@hursley.ibm.com>, Eliot Lear <lear@cisco.com>, discuss@apps.ietf.org
> > The question is, what are the applications-specific arguments > > against translated addresses? What are the brokenness conditions > > caused by translated addresses (and the associated statefulness)? > > FTP, SIP and DNS are three protocols which have problems. Or not really > SIP but anything which uses RTP and is launched by an SDP specification. > > Or, if you look at the situation when more than one host share one IP > address (not really address translation, but anyway, the normal way NAT > is implemented -- as 1:N mappings) and you have more than one host > listening for incoming connections, like an SMTP server. see: http://www.cs.utk.edu/~moore/what-nats-break.html and also section 1.1 of http://www.cs.utk.edu/~moore/I-D/draft-moore-nat-tolerance-recommendations-00.txt
Received on Monday, 2 December 2002 14:45:18 UTC