IPng:  ALE

Presented by Frank Solensky/FTP Software and Tony Li/cisco Systems


Frank Solensky

Biography:  Frank Solensky recently joined FTP Software as a project
manager and has been an IETF participant since 1989.  At the Vancouver
IETF in 1990, he presented an analysis of the growth of the Internet
that warned of the impending address assignment crisis.  Frank has been
reporting his projections of the size of the NSFNet Policy Routing
Database to the Big-Internet mailing list.


Figure 1 tracks the assignment of IP network numbers out of the Class B
and Class C address space.  They do not include the network numbers
which have been allocated to the service providers but not assigned to
any requesting organization, nor those which have been assigned but not
reported back to the NIC. The ``best-fit'' logistic trend lines are also
included for the data since mid-1992.

Figure 2 illustrates the growth in the NSFNet Policy Routing Database
which should be proportional to the size of the routing tables on the
backbone routers.  It also presents some estimates on the effect that
CIDR deployment will have on the same curves:  the middle curve assumes
route aggregation only at the campus level while the lowest one includes
the effect of aggregating routes at the provider level.

It should be noted that these curves are based on historic data and, as
a result, estimate the growth of the Internet only within the current
technology base.  Even if these estimates are off by a full order of
magnitude, however, we can be reasonably confident that the deployment
of CIDR gives us breathing room before IPng must be deployed.


Tony Li

Biography:  Tony Li has a Bachelor of Science degree in Mathematics from
Harvey Mudd College and a Ph.D. in Computer Science from the University
of Southern California.  He currently wrangles bits for cisco Systems,
Inc., specializing in high speed packet switching, interdomain routing
protocols and really ugly bugs.


The projections for address space utilization made in Houston are
updated here with slightly more data.  The projections made in Houston
appear to be consistent with what is happening.  There is, as of now,
insufficient data to further refine our projections.