[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Draft of updated WG charter
On 9-jan-04, at 23:30, Jay Ford wrote:
Path selection is only part of the multi-homing problem, & perhaps the
easier one. In my realm of experience, most end systems have only a
single path out, so there is no path selection required.
The nastier problem is source & destination address selection. In the
of multiple addresses per host, the host is forced to make the
the implications have to be handled in the rest of the network
filtering, routing policy...).
I'm not sure what you mean by "path" if you say that a host only has
one. I think for the purpose of discussing multiaddress multihoming, a
good definition of a path would be a combination of source and
The good part is that as soon as we implement mechanisms that allow
transport sessions to jump addresses, the address selection problem
isn't really a huge deal anymore: if you find yourself using
unfortunate addresses, you simply jump to something more suitable.
I'm not claiming that end systems should be completely stupid & the
should do everything. We might just disagree on where the
dividing line should be. I think address selection & path selection
shouldn't be on the end system side of the line, so an architecture
causes every end system in every multi-homed network to do those jobs
broken to me.
So what about having special boxes that sit between the hosts and the
routers and handle multihoming? Obviously anything that can be
implemented in an external box can also be integrated into a host when
desired, so unless the drawbacks of allowing this functionality to be
implemented in an external box are huge, this shouldn't take anything
away from the people who actually like their hosts to handle this