[p2p-sip] New I-D: The effect of NATs on P2P SIP Overlay Architecture

[philip_matthews at magma.ca (Philip Matthews)]

Eric Cooper and I have written an internet-draft that talks about
the effect of NATs on the architecture of a P2P SIP overlay network.

In brief, the draft begins with a list of assumptions on what
types of NATs exist in the network and where the peers are
located. We argue why we feel these assumptions are the
right ones. We then describes the consequences of these
assumptions on the P2P SIP overlay architecture.

Perhaps the most interesting assumption is that, in at least some
P2P SIP networks,  _ALL_ the peers may be located behind NATs.
We argue that, with the increasing proliferation of NATs today,
this is a very realistic possibility, especially in a smaller P2P SIP
network or in a P2P SIP network where a large proportion of
the peers are enterprise users.

The draft goes on to describe various ways an overlay network
can be constructed given these assumptions.

This draft is NOT trying advocate one particular overlay architecture.
Rather, it is trying to show the range of architectures available
given the set of assumptions, and give some measures by which
the various architectures can be compared. However, it does
conclude that there is really not a lot of choice -- given the
assumptions, there seems to be only one overlay architecture
that makes any real sense. This is a partial-mesh architecture
exhibiting a property we call "symmetric interest".

Note that the draft focuses ONLY on the structure of the overlay
network. We do NOT describe how searches for users and nodes
might be done in the network. This is an orthogonal question
which is left to other drafts. Our goal is simply to describe the
connection structure between peers in the overlay network.

We are very interested any any comments people might have,
either on our set of assumptions or on our analysis of possible
architectures that meet these assumptions.

The plain text version of the draft is available at
http://www.ietf.org/internet-drafts/draft-matthews-p2psip-nats-and- 
overlays-00.txt

There is also an HTML version which some people might find
easier to read. We hope to get this version up on a website shortly,
but for now you can e-mail either one of us for a copy.

Here is the abstract from the draft:
    This document discusses the constraints that NATs put on the  
possible
    overlay architectures of a P2P SIP system.  Given what seems to be a
    reasonable set of assumptions on where nodes are deployed and the
    kinds of NATs they are located behind, the document concludes that a
    structured partial-mesh overlay network exhibiting a property known
    as "symmetric interest" is the most reasonable overlay architecture.


- Philip