Vadim Antonov <avg@pluris.com> writes:

Sean Doran wrote:

...

As tli pointed out the top of the hierarchy is not arbitrary, it must be default free.

Sorry for the nitpicking, but this definition has at least two flaws:

a) there's a bi-partite backbone configuration, where each half has default pointing to the other half. Both do not have to carry full routes. (Of course, this scheme has problems with packets destined to the blue, or can be extended to more than two partitions).

In effect, this entails the synthesis by equivalent areas of a superior level into which each party can default. I touched on this briefly in my previous message. With variable length addressing this kind of joint level-n-plus-one synthesis is easy; the new area simply encompasses sufficient bits to distinguish each level-n/level-n-plus-one IS participating in it.

b) multihomed non-transit networks may want to be default-free and carry full routing to improve load sharing of outgoing traffic. Since they are non-transit, they cannot be considered "top of hierarchy".

Yes, I believe I also mentioned Yakov's "pull" (did you see his slides at the IETF (and NANOG?) with which he presented his push/pull definitions?) can be used to optimize routing when strict hierarchical routing is inefficient.

Bingo.

We are in sync, Vadim. Surprise surprise. Sean. P.S.:

Note that IXPs are not "top" of the hierarchy, but just aggregators for essentially point-to-point links between tier-1 backbones.

This is a good way of putting it. I will steal it and use it myself from time to time.