Experimentation in LISP
charter-ietf-explisp-01

The IAB arranged a workshop in October, 2006 to focus on routing and
addressing issues [0]. This workshop identified scalability problems
in the global routing system. The research community is working to map
the overall design space and understand the tradeoffs between the
different solutions. This work happens in the Routing Research Group
(RRG). The solutions discovered so far have interesting behaviours
that are not fully understood by mere desk analysis. For instance,
some of the solutions have packet reordering and delay implications
that may affect higher layers. Some have MTU implications for large
parts of the Internet. Most solutions require deployment of new nodes,
and the incentive models for deploying them are not entirely clear.

LISP and the LISP Alternative Topology mapping system (ALT) is one
solution in the overall design space, under the general category of
"map-and-encapsulate" mechanisms. The purpose of this BOF is to form
an Exploratory Group (RFC 5111) at the IETF. The group will host
discussions and documents necessary to perform experiments that help
the community understand the above behaviours better, by using LISP
and ALT as an example. The expected outputs are:

• document(s) that describe open issues where experimentation
  may be helpful,
• document(s) that describe planned experiments and results thereof,
  and
• experimental protocol specifications (exp, June 2009)

As an Exploratory Group, this group has a finite lifetime of 18
months. At the end of those 18 months, and depending on the outcome
of the experiments and the design work from the RRG, the group
may either be terminated or rechartered. The group shall also:

• clearly label its results as experimental
• avoid design discussions that are within the scope of the RRG,
• refrain from spending significant amount of time on the well
  understood parts of map-and-encapsulate mechanisms
• demonstrate commitment to focus on experiments before
  submitting any protocol specifications for publication
  as RFCs

This group is only focused on LISP and ALT. Some results may apply to
other designs as well, such as solutions involving a similar mapping
system but a different encapsulation scheme. But this is not
guaranteed. If other proposals surface from the research community
with equally interesting questions that would benefit from similar
experimentation, future groups can be created for that purpose, as
long as there appears to be sufficient interest in the community for
such work. The interest is demonstrated via XXX independent
implementation efforts.