TCP evaluation: Round table discussion

From WiL

Jump to: navigation, search

In November, Caltech will host a round table discussion on TCP evaluation.

1 Meeting details
2 Goals
3 See also
- 3.1 Visitor's information
  - 3.1.1 Accommodation and Transport
  - 3.1.2 Maps

[edit] Meeting details

Date: 8-9 November, 2007

Location: room 248, Lauritsen building #48, High Energy Physics dept, Caltech, Pasadena, CA, USA.
- See #Visitor's information below

Accommodation: To be announced (possibly the Sheraton, but not confirmed yet).

Who?: researchers interested in benchmarking internet TCP protocols!
- Live video-conferencing via EVO <http://evo.caltech.edu/evoGate/FAQ/> for those who can't come in person can still participate.
Some people that attended (in alphabetical order):
1. Lachlan Andrew, Caltech
2. Larry Dunn, Cisco Systems
3. Lars Eggert, Nokia
4. Sally Floyd, ICIR
5. Wang Gang, NEC China
6. Romaric Guillier, INRIA ENS-Lyon
7. Sangtae Ha, NCSU
8. Doug Leith, Hamilton Institute (by EVO)
9. Cesar Marcondes, UCLA
10. Saverio Mascolo, DEE Politecnico di Bari (by EVO)
11. Ihsan Ayyub Qazi, University of Pittsburgh (by EVO)
12. Tom Quetchenbach, Caltech
13. Injong Rhee, NCSU (by EVO)
14. Robert Shorten, Hamilton Institute (by EVO)
15. David Wei, Google (by EVO)
16. Jerome White, Caltech (tentative)

Overall Goal: To agree on some basic parameters and metrics for TCP evaluation and benchmarking.

[edit] Goals

Below is a list of parameters/measures to be discussed. There is a separate tentative list of parameters agreed on, and I'll eventually put together an agenda.

a set of bandwidths that are of interest, say 10, 155, 622, 2500 Mbps
a set of buffer sizes that are of interest, like BDP or 16384 packets
a set of distributions of RTT that are of interest
an agreed notion of "convergence time"
- e.g., "the average over period x is within y% of the
final average
an agreed notion of "time to converge to fairness"
- e.g., "the ratio of averages over period x is within y% of the final ratio"
- should this metric depend on the final ratio achieved?
an agreed notion of "intraflow variability"
- e.g., what timescales are of interest?
an agreed set of traffic models for background traffic
a measure of total link utilization
fluctuation in utilization due to fluctuation in background traffic.
What is the per-flow fair bandwidth share?

Let's sort out the "easy" easy topics above, and sort them out on the list before an interactive meeting, to save time for more debatable ones. It would be good if everyone can throw in some ideas for the next couple of weeks so that we can see issues are the hard ones.

It would be good to have a common set of scenarios which could be tested by simulation, emulation and real networks. Obviously, the simulation is the most flexible, and so it may have a larger set of tests, but we can at least simulate the emulated cases.

[edit] See also

[edit] Visitor's information

To ensure adequate catering and facilities, please contact Betta Dawson in advance if you have any dietary requirements, accommodation questions and for enrollment confirmation. Thank you.

[edit] Accommodation and Transport

Visitor's accommodation can be facilitated in advance via Betta Dawson (phone 626 395 6247) as soon as possible. Each attendee is responsible for paying their accommodation and shared rooms also can be booked in advance.