Re: Geographic v. topological address allocation

[Tony Li <tli () juniper net>]

> What I was trying to postulate, unsuccessfully, there is no such
> thing as an universal, optimal hierarchical addressing scheme. 

Sean,

It's pretty much obvious that the obvious definition of 'optimality' can
only be achieved if the topology is strictly hierarchical.  

Further, there is probably a 'maximal' addressing scheme for a given
topology.  But's only useful for the next 10 seconds that that topology
continues to exist.  ;-)

The other point that I think you're trying to make is that the maximal
hierarchy must follow the topology, and in some cases, this may actually
cause addressing divisions to not follow organizational boundaries.
Further, the organizations involved must be the ones to recognize this and
follow thru appropriately.

An interesting metric might be if the organization's min-cut-set bandwidth
is exceeded by its regional access bandwidth.

Tony