Internet Network Design Meeting
Westin Embassy Row Hotel
2100 Massachusetts Avenue, NW
Washington, DC 20008
November 27-28, 2007
Return to Meeting home page
FIND Working Meeting—Topics and structure
for breakout sessions
November 27/28, Washington DC
We have four breakout sessions:
Networking at the information layer
Moderator: Kevin Fall
Networking in times of disaster Moderator: Nick Feamster
Networks that provide high reliability and time-critical delivery Moderator: Dan
Networks designed to make management easier Moderator: Mark Allman
The purpose of the breakout sessions is to
carry out an exercise in which the group sketches an architecture for a network
to meet a specific set of requirements. The different breakout groups have been
given a different “target network”—a different set of requirements to meet. We
hope that by narrowing the set of requirements, rather than looking at a very
general target of “a future network for everything”, we can make this initial
exercise more approachable.
We are suggesting an approach that the breakout
group can take for this exercise—a process in five parts. While the groups are
free to modify the approach (how could we stop you?), this outline may help make
clear how we conceive the exercise.
The breakout has been split into a first part
on day 1, and the bulk of the work on day 2. It sometimes helps to take a first
step, and then take time to think, so we have set up this split format.
Step 1: Make sure there is convergence
in the room as to what the objective of the target network is. We have
offered a very brief description of the target, to help folks get into the
right breakout session. But we expect you will need to make this description
- Refine and elaborate the requirements
that the target network should meet.
- If there are alternative views as to
the right set of requirements for the target network, capture those
views and then converge on the specific set of requirements that this
group will explore on day 2. (Try to settle on a scope for the
discussion on day 2 that seems tractable.)
- Prepare a brief writeup of the set of
requirements, as the group has agreed, and any other viewpoints that
were raised in the group.
Prepare for day 2:
- Review the agenda for day 2. Is the
group ready to go?
- Does this group want to interact with
other groups? (If your breakout group wants to have a common discussion
with another group, try to arrange this with the other group moderator
on day 1.)
- Identify volunteers to prepare summary
slides and report back to the plenary group at the end of day 2, and
volunteers to help prepare a writeup of the breakout.
- Make a list of “lightning talks”.
We have proposed that each breakout have a
session of “lightning talks” by participants, in which they describe (in 3-5
minutes) the relevance of their work (FIND funded or other) to the target
set of requirements. (This talk might be based on the poster that they
prepare for the evening session of the first day.) The lightning talks are
suggested as an early item on the second day. On the first day, the group
should make a list of lightning talks that they will hear.
If you have time, start on the day 2 agenda: start to catalog related
Step 2: Identify and discuss possible
approaches to building the target network.
- Lightning talks: review relevant
approaches from inside the FIND community.
- Related work: catalog relevant work
being done elsewhere.
Step 3: Try fitting the various
- How do the identified approaches (from
step 2) fit together? Are they consistent or at odds with each other?
- What is missing? Where are
- How do these requirements and the
known methods to achieve them influence or interact (or interfere with)
other requirements (e.g. overall security, manageability, emerging
- Is it possible to sketch an overall
design? If so, do so.
Step 4: Draw some conclusions and think
about going forward.
- Are there non-technical barriers to
achieving these goals?
- Are there identifiable options for
mitigating these barriers?
- What other researchers (or research
disciplines) should be engaged as part of addressing these requirements?
- How close are we to meeting the
requirements? What is the level of risk in trying to achieve them?
- Are there folks in the room who want
to continue this discussion? Can you identify volunteers to improve and
elaborate what you have written in the breakout? Can you set a target
(e.g. a month) for an improved writeup?
Step 5: Capture what you have learned.
- For the group/NSF: summarize your
discussion, using this outline as a framework. We have provided a slide
template, which may be helpful in getting started. We also encourage a
text-based writeup of the important discussion and conclusion, using
what your note-takers provide you.
- For GSC: The GENI Science Council is
collecting examples of important outcomes and exciting research that
will lead to those outcomes. The GSC also wants to identify any specific
requirements that GENI must meet in order to perform the research. If
your breakout has made progress, what you have done could be such an
example. Please prepare a short writeup that we can convey to the GSC,
- What is the impact of meeting this
set of requirements?
- What is the important science and
how can it be explored?
- If these ideas were to be
evaluated/demonstrated on a research facility like GENI, are there
any special requirements for that facility?
If there is time...
- We posed four challenges for the
breakouts—four different sets of target networks with different
requirements. Are there others we could explore?
- Are there topical workshops we should
consider organizing, in order to carry the research forward?
Return to the plenary prepared to give a
10-15 minute report at 2:00 PM.
We hope this was fun and effective. Tell us.