Charles Hannum writes:
> In a multitasking system, this is a different issue.  What happens if
> some user decides to read the routing table 1000000 times?

What happens if some user starts a process that jams data out the
ethernet as fast and hard as possible down a UDP port?

> Okay; suppose I'm convinced of this.  What would you propose doing,
> that allows reading the tables from crash dumps and across the network
> (both of which are important for debugging)?

I've got to note that I don't think that tools for reading network
status across the network are important enough to change our
architecture for local status reads.

Perry