If you are concerned about fixing it, and the outage you will or
are causing to your customers, perhaps you ought to think about
discussing why this is so with either your architect/engineering
group or perhaps with your upper management/finance group.
One would hope that the major backbone players would at least
engineer thier backbone to survive or be able to recover quickly
from a single failure event for equipment that has shown that
tendancy. We all know routers don't have the greatest
reliability standards.
Backbones who have overloaded routers either are working under
a very risky business model, or they have not desinged and
engineered thier network correctly? I suppose in some small
cases with lower performance routers, something like this
could possibly sneak up on you, however, in the statistically
aggregated traffic models of a larger backbone, this shouldn't
be able to sneak up on you.
Just my plug for intelligent backbone engineering. I learned
the hard way.
Chris A. Icide
Nap.Net, L.L.C.