Keith Moore wrote:
It seems to me that the general problem is not multiple interfaces, but
multiple addresses per host. It doesn't matter (much) whether those
addresses result from multiple physical interfaces, a combination of
physical and virtual network interfaces, multiple prefixes being
advertised on the network attached to any particular interface, or even
a mixture of v4 and v6.
Not quite right, IMO. I think that the problem stems from having
multiple _prefixes_ per host, because IP hosts don't generally have a
way to associate configuration information with a particular prefix, nor
do they have good ways of telling which prefix is most appropriate for
reaching a certain destination. Hosts that implement all of the IETF
protocols (and nothing that tries to do better) currently use
longest-match, but that doesn't always work well in the "real world".
So that might have some impact on the name, particularly if you want to
attract the breadth of interests whom this affects. Something like
Hosts Addressed Multiply (HAM), perhaps
As I indicated in my previous message, I don't see the point of changing
the name, especially not to reflect a concept that this is a multiple
addresses problem when I don't think that is actual the case. One of
the things that this group can/should do is try to figure out where the
problem lies. We know that some multi-interface hosts work just fine --
where all of their interfaces are under a signal administrative control
and/or connect to all of the services that a host wants to use. But,
there are other cases (corporate VPN, WLAN/cell devices, etc.) where our
current standards don't work well. Figuring out exactly where things
break down is part of this effort.
Margaret
_______________________________________________
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf