this is why some of us do it right the first time. the national isp
that i helped architect recently has not a single point of failure in
the entire infrastructure. in fact, we scaled it such that if we had
catastrophic losses in any given machine pool (this usually means up
to two machine failures for a given service) we could still handle
full capacity. ah, i love it when a plan comes together... <grin>
of course, if a class five earthquake hits san jose, we might
experience a temporary network outage...
jrk
-- Joseph R. Kiniry Graduate Student, California Institute of Technology