[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Architectural approaches to multi6

Ran,


Does anyone else have any other architectural approaches
(besides the two mentioned so far) that the WG should
consider ??

Existing architectural approaches seem to be:
    1: identity/location separation; with the locator
        following the actual network topology & the
        identifier being quite independent of topology.
    2: tunneling to create a virtual topology
        that matches the addressing

Well, a number of people have suggested to work on
addresses.  The various PI schemes may be unworkable
given the current hierarchical addressing and routing
structure; I don't know.  But I think that there
might be some benefits to be found if we took a deeper
look at the interplay between addressing and routing.

I am definitely not an expert in the area, but I have
heard people speaking about clustering based addressing
etc. instead of hierarchically assigned addresses.
Assigning the addresses somewhat dynamically based on
the real network topology, perhaps combined with an
identity/location separation, may be something worth
considering in the long run.

But that is probably too far fetched for this WG.

--Pekka Nikander