North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: multi-homing fixes

  • From: Sean M. Doran
  • Date: Wed Aug 29 08:39:12 2001

Having lived for a while now in a small country, whose content for the
most part is encrypted into a language only a few million people can
read, I can assure you that there is nontheless alot of traffic flowing
into the eyeballs of people living there (even though it's in other
languages including unencrypted American), and the vast majority of
that flow is paid for by the people living in that small country, 
whether they're paying-to-see or paying-to-be-seen.   The heavy
politics are all in who gets what piece of that cash flow.

CNN has a local organization maintaining translated and locally-written
material, for what it's worth.   And anyway, with Akamizing and AOLephanting
and so forth going on all over the place, they're a bad example for many
other reasons.  (Not least of which is that I get better news coverage

But who would care if Bill's Bait Shop in California stopped being
able to talk to a local national provider?  And who would pay to fix
that connectivity, if it were fixable only through the application of money?

Who knows?  (I'll take yer 50 cents though, Bill).

Anyway, the whole idea of universal connectivity is a complete myth.
Even when the connectivity is advertised at any given moment, it's
not necessarily there.   Multihoming in practice quickly reveals
that "full routes" are not the same size and do not lead to the
same goodput, as noted in part by Valdis.