On Wed, 9 Jun 2004, Masataka Ohta wrote:
> Pekka Savola;
>
> > There is only marginal value in
> > ND using multicast (compared to broadcast), as the typical number of
> > nodes per link is so low that broadcast storms are not a real problem.
>
> >>The solution has been never bother to have standard way of
> >>address resolution.
>
> > I have to disagree here. Just because broadcast/multicast (ND) isn't
> > optimized for WLAN (or CSMA/CA in general, I guess), that doesn't
> > nullify the usefulness of ND for all the other media for which this is
> > not a problem.
>
> It has been wrong to try to make ND "useful" for all the useless
> purposes, some of which is already recognized useless or
> harmful (for example, at first, as reasoning not to use ARP, it
> was proposed to send unicast data packet over link multicast,
> which was expected to eliminate the need for a pending queue).
> Now, it is more and more widely recognized that stateless
> autoconfiguration, which was invented as reasoning not to
> use DHCP, is not as useful as expected.
I think it is very useful. The DHCP is very heavy in some environment.
Sorry Masataka, but you are mixing autoconfiguraiton and neighbor
discovery. They are different beasts, they are using the some
infrastructure, that is called ICMPv6.
>
> The sane thing to do is to narrow the scope as much as possible
> and to remove features as much as possible.
>
> For example, MIPv6 suffer, because it, seemingly, recognized ND
> available on all the link types and developped protocol. Then,
> it was found that ND timing values makes MIPv6 not useful for
> quickly moving hosts. It is wrong to have standard timings
> ignoring link specific properties. And, then, there is no
> reason to have standard protocols ignoring link specific
> properties.
>
> If we concentrate on Ethernet and address resolution, there is
> no reason not to use ARP.
There is a reason to use ND e.g. on point to point to links you can use
ND, while ARP is useless. ND is very good at discovering neighbor in a
unified way. I think there is no reason kill ND, but make it optimised for
CSMA/CA type network.
Regards,
Janos Mohacsi
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[EMAIL PROTECTED]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
