wiki:Gec5ClusterCAgenda

Version 3 (modified by ricci@cs.utah.edu, 10 years ago) (diff)

--

Cluster C Meeting Agenda for GEC5

The GPO has given us two timeslots to meet as a cluster at the next GEC. Below are draft agendas for those meetings. Feedback and suggestions welcome: send them to Robert Ricci

Note: Though these are meetings for Cluster C, anyone is welcome to attend. The meeting on Tuesday is more likely to be of interest to those outside of our cluster, as the Monday time will be used on demos, which can also be viewed at the demo session later in the evening. However, if you're giving a demo in the evening session, and thus won't be able to walk around to see other demos, you might consider stopping by our meeting!

Please note two things:

1) For those of you who have signed up to do demos, there's a strong chance that our cluster meeting will happen in a meeting room, not the demo room. The meeting room is supposed to have a projector and WiFi. Will this be okay for everyone?

2) I have a slot in the agenda for the second meeting for projects to give (short) status reports, but only one group signed up so far. If anyone else would like to give a report, please let me know.

Cluster Meeting: Monday, July 20, 1:00 - 5:00

Because it's hard for those giving demos in the demo session to go around and see others' demos, and because it's particularly important for cluster members to see each others' demos, we'll be using this time as a sort of "dress rehearsal": a chance to present our demos, one at a time, to each other. The set of projects that have agreed to present are:

ProtoGENI: Robert Ricci (Utah)
Digital Object Architecture: Giridhar Manepalli (CNRI)
DTunnels and BGPMux: Nick Feamster (Georgia Tech)
Million-node GENI: Justin Cappos (and others?) (Washington)
Measurement System: Charles Thomas (Wisconsin-Madison)
Programmable Edge Node: Yan Luo and Tim Ficarra (UMass Lowell)
TIED: Ted Faber (ISI)

I will not be putting specific time limits on these demos, since I expect we will not use the entire four hour slot that's been alloted to us. However, do keep in mind that people will be watching several demos during this session, and that there will be a demo session afterwards, so please try not to go too long.

Logistics: Depending on the number of people who choose to attend and the sizes of the spaces, we may hold this session either in the "cluster meeting room" or the demo room. We will have access a projector and wireless Internet in the cluster room. It's unknown at this point how early all of the equipment in the demo room will be set up.

Cluster Meeting: Tuesday, July 20, 1:00 - 3:00

This meeting will give us an opportunity to discuss some of the technical challenges facing our cluster. Most topics within this meeting will be group discussions, kicked off by a short presentation.

A. Introductions and Status Reports

This will be a chance for any projects that wish to introduce themselves to the cluster or to give status reports. Any other group that would like to sign up for this agenda item are welcome to; presentations may be as formal or informal as you'd like.

  • ETRI Korea has proposed to join this cluster in Spiral 2; they will be giving a status update

B. Control Framework Plans and Priorities

This session will be a chance for the ProtoGENI project to present its plans for Spiral 2, with particular emphasis on how this will affect the other projects that rely on it as a common control framework. We hope this will stir up some feedback, helping us to alter our priorities to match what's needed by others in the cluster, and to help set priorities for the cluster as a whole.

C. Design Discussion: Integrating Million-Node GENI

The Million-Node GENI project from the University of Washington has joined Cluster C, and presents a particularly interesting integration challenge. In many ways, it has semantics that do not mesh with those described in many of the GENI design documents and on the ProtoGENI wiki. In many cases, this seems to be due to the fact that this project aims to control a very large set of resources (millions of nodes) owned by many different owners (potentially tens or hundreds of thousands.) The semantics that make sense for such a loose collection of resources are different than those in the cluster's current model, and working with the Million-Node GENI project presents an opportunity to expand our control framework to encompass such resources. This discussion will center on the ways in which the design of our control framework can be informed by the experiences of the Million-Node GENI project.

D. Backbone Discussion

In partnership with Internet2 and HP, the ProtoGENI project is in the process of rolling out a nationwide backbone. Connectivity on this backbone is planned for several GENI projects and campuses. This portion of the meeting will give updates on the design of the backbone, status reports, and ideas for connecting to projects and campuses.

Attachments (4)