Re: Alternative to BGP-4 for multihoming?

[Andrew Bender <abender () tns-inc com>]

Gary E. Miller wrote:
> And I sure hope that HOP count is not used, Singapore has a 
> mixture of very fast and very slow links to foreign countries 
> so hop count is a useless measure of "closeness".

Quite possibly, if used by itself. Dr. Jacobsen's method
(http://www.caida.org/Pathchar) is a proven strategy that could
provide some relief here, but may be impractical interactively.
Then again, isn't any interactive measurement going to be 
impractical?

Andrew Bender
tns-inc.com