Brian E Carpenter <[EMAIL PROTECTED]> wrote:
|Dan Lanciani wrote:
|>
|> "Tony Hain" <[EMAIL PROTECTED]> wrote:
|>
|> |That said, there are multiple parts to the isolation issue, and even though
|> |most NAT implementations combine them, the discussion will be more
|> |constructive to keep them separate.
|>
|> I used to believe this, but I recently came to the realization that isolation
|> from provider address policy is a single issue.
|
|Perhaps, but it is not an issue to which NAT is the only answer.
I never implied that it was. There are many simple solutions including PI
space handled at the routing level, source routing in any of several forms
including locator/identifier separation, and overlay networks. Unfortunately,
none of those solutions seems to be able to make it past the politics of
provider-controlled address allocation. The required technical work to support
them never happens or (in the case of overlay networks) technical tweaks are
implemented to prevent end users from implementing it themselves.
I see two classes on solution:
1) Solutions that require that we make some change to the protocol. These
would include the approaches I listed above. If you know of such a solution
that isn't DOA (i.e., that is even allowed to be discussed on this list) then
please describe it.
2) Solutions that end users can implement without change to the protocol. I
had hopes here for overlay networks, but these still require access to address
space that is not burdened with punitive semantics like default router rules
and DNS prohibitions. Even with such access it would be difficult to establish
critical mass and turn the overlay network into the primary network. If you
know of a solution that the market can provide without protocol support (other
than NAT) I'd love to hear about it.
|In IPv6 we have
|enough address space to solve it without NAT.
Address space was never the problem. Portable/global/routable address space
available directly to end users (or rather lack of it) is. You can give the
service providers all the address bits in the universe, but that won't solve
the problem.
People have been claiming for years that IPv6 can deliver the same (or better)
functionality as NAT, but more cleanly. Recent comments (including some private
email) suggest that this claim carries the unstated assumption that the only
functionality provided by NAT (or at least the only functionality that end users
should be allowed) is conservation of the total address pool. Assuming folks
really believe this, I think the disconnect with the real world is so great as
to make progress virtually impossible.
Dan Lanciani
[EMAIL PROTECTED]
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[EMAIL PROTECTED]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
