Folks,
Regarding Martin's question if the situation when the outgoing traffic
passes a different path than the incoming traffic is real:
Yup, it's real. Please see the Table 1 in the paper entitled ``Optimal
Choice of Peers based on BGP Information'', which is cited in
draft-dulinski-alto-inter-alto-protocol-00 and
draft-dulinski-alto-inter-problem-statement-00, and is available from [1].
Regarding Jan Seedorf's comment that they have made experiments and that
they confirm that the inter-ALTO aka. server-to-server communication helps:
We have also confirmed the need of the inter-ALTO communication. Please
see again [1,2] (where the AS-path length is taken into account) and the
paper entitled ``Cost-driven Peer Rating Algorithm'' [3] (where the
traffic costs are taken into account).
[1] Free source:
http://www.kt.agh.edu.pl/~cholda/wp-content/uploads/2011/02/ICC2010_1.pdf
[2] Surely permanent, but paid source:
http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5502269
[3] Free source:
http://www.kt.agh.edu.pl/~wydrych/wp-content/uploads/2011/01/Cost-driven-Peer-Rating-Algorithm.pdf
(permanent link will be provided by IEEE after ICC11)
Best Regards,
Piotr 'GhosT' Wydrych
--
Piotr 'GhosT' Wydrych .. xmpp:wydrych//agh.edu.pl .. http://wydrych.net/
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
_______________________________________________
alto mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/alto
