On 8/29/07, Arnaud Ebalard <[EMAIL PROTECTED]> wrote:
> Hi,
>
> "Christopher Morrow" <[EMAIL PROTECTED]> writes:
>
> > and 80x magnification is a problem... but in this case (and in many
> > other cases of same) there are simple solutions that could/should be
> > used. Solutions which don't require base protocol changes.
>
> Can you describe those _simple_ solutions and compare them with the
> deprecation? ;-)
>

Sure, if you see a problem related to the ping/pong between 2 devices
simply filter that traffic. The sky is not, in point of fact, falling.
There is no difference between this sort of security problem and
smurf-attacks or syn-floods. (speaking as someone that's had to
protect thousands of customers and a sort-of-large network for 8+
years from these sorts of issues).

> IMHO, the real problem is that RH0 should not have been in the IPv6
> specifications (1883 _and_ 2460) and the associated time/space could
> have been used, as you wrote, for something more useful.
>

Sure, the original idea was probably (I've not read the history) to
have feature parity as much as reasonably possible. I suspect that
when 1883 and 2460 were written source-routing in ipv4 was still
somewhat useful. Today, for a variety of reasons, it's become mostly
not-usable except in specific applications.

> locator/id split is actually one of such things, but having it directly
> in IP has pros and cons. Keeping IPv6 simple and efficient allows
> building on top of it (like MIPv6 does) with some insurance on resulting
> stability.
>

Certainly, doing so scalably (like making all of ipv6 MIPv6 instead of
a special class of MIPv6 capable endopints) would have been a much
better long term solution :( Essentially, as a number of presentations
I've given or [EMAIL PROTECTED] has given, the world IPv6 was designed for
didn't appear. The assumptions inherent in the design (the largest
mis-assumption being a small number of large providers with aggregate
routing and addressing) have driven us to a wierd place...

thanks,
Chris

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to