On Tue, 2009-02-03 at 17:03 -1000, John Zwiebel wrote:
Much of the aggregation happens just by deploying LISP/ALT since the
route advertised into the ALT by the xTR is already aggregated.
But then there is no need for a heirarchy right? We could get the
same
benefits by having a set of public-use routers that can return map
replies, peered in a flat format.
The motivation behind a heirarchical structure for the ALT-LAT was
aggressive aggregation, which leads to the issues articulated by
Brian,
where globally-spread sites have to send all map requests through one
geographical point, leading to some major stretch.
Dan, this is very simple, we do and will have a hierarchical
structure. And when there are allocations they have to be reported
within an aggregation boundary.
That's it!
Dino
However John, you're right that IF edge sites are willing to use
LISP's
TE methods, they won't have to prefix split for TE. This means 1
prefix
per edge site, which is better than today. But then, as I said, we
can
get the same benefit without an ALT tree.
Dan Jen
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
