ietf
[Top] [All Lists]

Re: solution to NAT and multihoming

2001-01-26 11:30:02
Keith, Ed, others...

I have been following this entire line of discussion with some
amusement and some frustration. I would like to share a couple of
humble thoughts on this subject from my own perspective.

- yes, NAT in general restricts the applications and/or protocols that
can be accessed by users behind the NAT.

- yes, having NAT devices deployed will impede development of new
protocols and applications that rely on embedding IP addresses.

- yes, NAT can be cumbersome in its sustained management as sys admins
must punch holes through their various NAT devices to allow a
particular application/protocol through.

- yes, NAT does violate the global addressibility of Internet hosts.

- yes, we could eliminate NAT by giving everyone a globally unique IP
address.

- no, not everyone wants to run every conceivable application/protocol
to their client machines, they are happy with the subset they chose.

- no, protocol developers cannot go off developing new protocols that
do not consider implications with NAT deployment.

- no, not all NAT implementations prevent the use of punch-throughs
allowing unique or custom protocols to still work.

- no, not everyone wants to participate in the great global address
space of the Internet, they just want to access Internet-connected
devices.

- no, as a mere mortal user, I cannot always obtain real IP addresses
as the ISPs claim they must justify IP address assignment and hold them
close to their vest pocket.

Considering my own company and the plethora of IP-addressed devices,
and yes we sit behind NAT and yes I can do nearly all applications and
protocols as one can who is not behind a NAT, many of these devices are
lab tools that only need connectivity back to an engineer's desk or a
shared printer. I don't really need access to a JTAG emulator pod from
across the ocean, it just doesn't make sense for my purposes.

Given the argument that NAT restricts the available applications and/or
protocols, a potential buyer of the device must then choose the one
that meets his or her requirements. Since the more restrictive devices
will most likely be purchased less and less, and the "better" devices
will prevail, will it not be expected that the NAT implementations will
improve over time in a manner similar to a farmer improving his crops
year after year by only planting the hardiest varieties?

So when it comes to buying NAT devices, "buyer beware" should be the
mantra of the day.

And now the question(s) of the day:

What is the solution that can be deployed today or in the next 6 months
that will replace the function of NATs in the IPv4 Internet? What about
in the next year? 2 years?

Respectfully,

Kevin Farley




__________________________________________________
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices. 
http://auctions.yahoo.com/