Vulnerbilities of Interconnection

Sean Donelan sean at donelan.com
Fri Sep 6 00:58:47 UTC 2002


On Thu, 5 Sep 2002 sgorman1 at gmu.edu wrote:
> I completely agree with statement.  It is not a matter of wanting to
> know where the importants hubs are - we have a pretty good handle on
> that, but what the impacts would be of a hub loss from an operational

I don't know about that statement.  There may be some super-secret list
of the most critical locations which I'm not privy too.  However most of
the lists I've seen suffer from brand-name bias.  They have well known
locations, which is not necessarily the same as the critical locations.
And as others have pointed out, a lot depends on what services you define
to be critical.

The Internet is a non-linear system.  The impact of a loss of a hub or
even many hubs doesn't depend so much on the hub, but on what else is
happening in the network.  The Internet is "failing" all the time, all
day and all night.  Why don't we notice most outages?  What causes a
disruption to grow large enough to impact lots of people?






More information about the NANOG mailing list