Re: on NAT-PT

[Margaret Wasserman <mrw@windriver.com>]

> As a WG contributor I disagree.
> I think both NAT-PT and SIIT should be moved to historic or at least
> to a status where their deployment is not recommended.
Also as an individual contributor...

I agree with Erik.  I haven't seen a case yet where NAT-PT cannot be
replaced with a less damaging and simpler mechanism (or combination
of mechanisms), such as dual-stack hosts and tunneling.

> I assert that 90-99% of the cases where they are used one can instead
> use dual-stack hosts (and routers) with private IPv4 addresses and
> a good old IPv4 NAT.
> [...]
> So I think the preferred scheme for this type of interoperability should
> be what we already have - IPv4 private addresses and IPv4 NATs.
> With IPv6 end-to-end communication when communicating with IPv6 peers.
Exactly.  Any node that needs to make a connection to an IPv4-only
node or service must include an IPv4 stack.  Any node that needs to
connect to an IPv6-only node or service must include an IPv6 stack.
Nodes that need to do both must be dual-stack.  It just make sense.

NAT-PT also has some problems that would be a serious limitation for
almost any deployment, such as the inability for two nodes on the
IPv6-only side of a NAT-PT box to establish an IPv4 connection to
each other.

> Of course, there is 1-10% of cases when this is unattractive e.g. due
> to the need to route IPv4 inside the domain and it would make sense for the
> WG to try to understand that space better. In some of those cases
> I think the dual routing will be simpler than having a new type of NAT device
> to deal with. In other cases it might be sufficient to have proxy 
> functionality
> at the v4/v6 boundary and no NAT (e.g. SIP, HTTP, SMTP)
There are also other alternatives that would not require IPv4
routing within the domain.  For example, a dual stack host running
in an IPv6-only infrastructure could set-up an IPv4 in IPv6 tunnel to the
a dual-stack router on the edge of the IPv6-only network.  The IPv4 address
for the dual-stack host could be dynamically allocated/leased, allowing a
large number of nodes that occasionally use IPv4 to share a smaller
number of IPv4 addresses.  DSTM offers one example of this type of
mechanism.

The proxy alternative may make sense for the 3GPP IMS case, as this
is a special case where only certain services are required to be
IPv6-only.  IMS-capable handsets may still be capable of opening
IPv4 PDP contexts for communication with non-IMS IPv4 nodes and
services.

I don't think that we've explored these alternatives enough to know
which would make the most sense for the 3GPP IMS situation, and I
don't think that we should settle on NAT-PT as the solution for that
scenario until we have actively looked for alternatives that avoid the
problems of NAT-PT.

Margaret