History of E2Epi

One of the founding premises for Internet2 is the premise that Quality of Service (QoS) will solve reliability problems experienced through the commercial internet. It is presumed that, until QoS is going, bandwidth overload will solve problems.

Eventually, there is a realization that bandwidth alone doesn't solve all problems. Matt Mathis of PSC posits a Wizard Gap; those with technical know-how can nearly always use a system to full or near full potential while the average user is experiencing end-to-end (E2E) problems or an insignificant increase in speed.

2000

01 November 2002 - At the Fall Member Meeting, Laurie Burns and Steve Corbató held discussions about the need for an initiative to solve the E2E performance problem.

Planning for January 2001 workshop in Ann Arbor to gather experts to brainstorm (44 invitees 14 staff, 30 members). Steve Corbató and Laurie Burns outlined an agenda for the meeting and called for white papers from participants to frame the scope of the discussion on the goals of the initiative.

2001

9 January 2001 - Workshop happens in Ann Arbor. No overarching consensus but many ideas. Decision to create a design team. After the meeting, created an executive summary for public review; white papers were made available through the web (see December).

12 January 2001 Small group tasked to develop a paper for discussion:

28 January 2001 - Guy Almes gave a presentation on the E2Epi at the Joint Techs workshop in Hawaii.

29 January 2001 - Small group meets at Joint Techs (Hawaii). Working session emphasized brainstorming. Divided work into 4 areas (applications, host/OS, network, and operations) with PERT and communications identified as goals.

15 February 2001 - Divided up team by areas -- Ted Hanss headed Applications, Basil Irwin headed Host/OS, Matt Zekauskas headed Networks and Russ Hobby headed Operations. The teams were to write up sections and Russ would integrate their material into a final document.

March

07 March 2001 - At the Spring Member Meeting, held in DC, George Brett (NLANR) presented a report on the work of the E2Epi design team.

09 March 2001 - During a session at the Spring Member Meeting on E2E performance, five presentations were made:

For the Closing Plenary session, Laurie Burns described the E2Epi; she solicited member involvement in the Initiative and comments on the Fat Pipes paper.

Search begun for Initiative Director.

Search for Initiative Director continued. Cheryl Munn-Fremon selected and hired.

30 May 2001 - A Campus Workshop was held at Georgia Tech (in Atlanta) focusing on engineering and E2E performance. Guy Almes, Matt Zekauskas, and Russ Hobby ran a workshop to discuss: What are the E2E performance problems on this campus and how can Internet2 help solve them? What is GA Tech doing to fix them? What works/doesn't work?

04 June 2001 - Cheryl Munn-Fremon begins work as Director of E2Epi. Pulls together team (Russ Hobby, George Brett of NLANR, and Terri Saarinen of UM).

07 June 2001 - Russ Hobby begins work with the E2Epi team as the Chief Technical Architect.

11 June 2001 - Terri Saarinen begins work as Program Assistant for E2Epi.

16 July 2001 - The E2Epi team met with some members of the Internet2 Engineering team in Ann Arbor. Russ Hobby, Cheryl Munn-Fremon, Stanislav Shalunov, Ben Teitlebaum, and Matt Zekauskas discussed various E2E performance-related projects currently under development (including Stethascope, Detective, Reflector, and Cake Boxes).

24-25 July 2001 - At the Web100 Workshop in Boulder, Colorado, attended by Cheryl Munn-Fremon and Matt Zekauskas, Matt gave a presentation on Measurement and E2E.

01 August 2001 - George Brett joins E2Epi from NLANR as the Chief Information Architect.

02-05 October 2001 - Virtual Fall Member Meeting (because of September 11 attacks). Cheryl Munn-Fremon gave a presentation on E2E. Russ Hobby gave a presentation on performance measurement.

10 October 2001 - Matt Zekauskas, Guy Almes, and Russ Hobby attended the AARnet conference in Australia. Matt gave a presentation on E2E performance; the AARnet group expressed an interest in working with Internet2.

26 October 2001 - Cheryl Munn-Fremon gave a presentation at the National Broadband Summit.

12 November 2001 - George Brett presented a report on the Internet2 E2Epi at CUDI.

14 November 2001 - At the SC2001 conference, Russ Hobby gave an E2E performance workshop presentation.

29 November 2001 - Russ Hobby presented an overview of E2Epi for the AllStaff meeting in Ann Arbor.

2002

30-31 January 2002 - At the Joint Techs Workshop in Tempe, Arizona, E2Epi hosted a Measurement Workshop. Presentations, papers, list of attendees, and summary information is available.

26 February 2002 - Eric Boyd was interviewed for the position of Performance Engineering Consultant.

12 March 2002 - Eric Boyd begins work with the E2Epi team as the Performance Engineering Consultant.

12 March 2002 - Cheryl Munn-Fremon and Eric Boyd visited ITEC-Ohio to establish a working relationship and ensure that work was divided in such a way as to ensure no overlap of effort. At the meeting, the details of working with Prasad Calyum were finalized.

01 April 2002 - Russ Hobby facilitated a conversation about measurement data coordination. Participants included representatives for the Quilt, Abilene, NOC, and the Internet2 Engineering team.

01 April 2002 - Russ Hobby facilitated a conversation on toolkits and GUIs to determine what everyone wants in a user interface. Participants included representatives from DAST and the Quilt; DAST agreed to develop the GUI.

10 April 2002 - Russ Hobby gave a presentation on the E2Epi at a Campus Workshop at Texas A&M (Houston).

15 April 2002 - The Windows on the Future Conference at the Ohio State University campus was a joint presentation between ITEC-Ohio, ADEC (American Distance Education Consortium) and Internet2.

15 April 2002 - Russ Hobby gave a presentationin Copenhagen at the NORDUnet conference on E2E.

25 April 2002 - A brainstorming meeting was held in Armonk, NY; participants included Eric Boyd, Russ Hobby, and members of the Engineering Team. This meeting generated ideas that lead to the development of the Performance Improvement Performance Evaluation System (piPEs).

07 May 2002 - At the Spring Member Meeting, Russ Hobby gave an overview of E2Epi for the Internet101 Tutorial session.

20-21 May 2002 - E2Epi hired 3 interns (2 from UMI, 1 from UTX) for the summer to assist in piPEs development.

June

04 June 2002 - E2Epi submitted a proposal to NSF for the development of piPEs.

05-06 June 2002 - Cheryl Munn-Fremon and Eric Boyd visited NC-ITEC to discuss coordination of efforts.

13 June 2002 - Eric Boyd met with representatives from DAST (in Ann Arbor) to discuss development of the GUI (see 01 April 2002).

18 June 2002 - A draft of the piPEs paper was posted on the E2Epi home page for comment.

July

25 July 2002 - George Brett gave a presentation on piPEs, including the current status, at the Global Grid Forum (GGF5) meeting in Edinburgh, Scotland.

30 July 2002 - At the Joint Techs meeting in Boulder CO, Eric Boyd first gave an update of the E2Epi and, at a BoF session later in the day, presented information on piPEs.

August

06 August 2002 - Eric Boyd presented an E2Epi update at the Internet2 Application Steering Committee Meeting.

19 August 2002 - At the Chinese-American Networking Symposium, Eric Boyd gave an E2Epi update presentation via teleconference.

14 August 2002 - Interns Clyde Shih, Jooyung Jun, and Ramanuja Vedantham finished work on two documents related to the Scheduler and the Testing Engine.

04 September 2002 - Eric Boyd gave an E2Epi update at the Internet2-SWITCH meeting.

07 October 2002 - Susan Evett began work with the E2Epi team as Technical Writer/Editor.

27 October 2002 - At the Fall Member Meeting in Los Angeles, members of the E2Epi team gave various presentations. Eric Boyd and Warren Matthews (SLAC) hosted a meeting of the piPEfitters, a group dedicated to the development and implementation of a performance measurement and evaluation system for Abilene. This was a closed meeting; participants included those who had volunteered to work on specific parts of the piPEs project or were involved in related work. At a separate, concurrent session, Russ Hobby gave an update on piPEs to the High Energy Nuclear Physics (HENP) working group.

28 October 2002 - Russ Hobby presented information on E2Epi and the piPEs project during the Internet2 101 Tutorial session. At a separate session, Eric Boyd, Russ Hobby, and George Brett provided an update on the E2Epi progress. George Brett and Terri Saarinen facilitated the Peer-to-Peer (P2P) working group meeting; work in progress was discussed and decisions about the 2003 meeting were made.

29 October 2002 - Eric Boyd hosted a BoF on the E2Epi, at which he gave a presentation on network engineering, focusing on the piPEs project. Warren Matthews (SLAC) gave a presentation on IEPM and MAGGIE; Rich Carlson (ANL) presented information on developing Web100-based network configuration and management tools. Jim Ferguson (NLANR/DAST) gave an update on the Network Performance “Advisor” project; Prasad Calyum demonstrated the Beacon tool.

30 October 2002 - Russ Hobby chaired a meeting of the E2Epi Technical Advisory Group (TAG). Matt Zekauskas chaired a meeting of the Measurement Working Group; John Moore (NC ITEC) provided an update on tests they have been running, Loki Jorgenson of jaalaM Technologies described how their product can help users quickly find the source of an E2E problem, and Matt gave an update on the Abilene Measurement Infrastructure (AMI). George Brett and Frank Connolly (ConseQ Associates) hosted a BoF on an National Library of Medicine (NLM) pilot study to assess user satisfaction with the internet.

11 November 2002 - At the American Medical Informatics Association (AMIA) Conference, Russ Hobby gave an overview presentation on the topic of E2E performance.

18-21 November 2002 At Super Computing 2002, Prasad Calyum gave a demo of Beacon and Jeff Boote gave a demo of the Abilene One-Way Active Measurement Protocol (OWAMP). Eric Boyd talked with Bob Grossman (UIL) about using his tool in piPES. Eric also had a discussion with Bill Wing from DOE (Oakridge) regarding his research on Layer 2 performance analysis; they set up a meeting for December 12 with John Streck and John Moore of NC-ITEC about Bill's work and continuing it on Abilene with Spirent gear. Eric also held a discussion with Chris Herman, Matt Zekauskas, Christian Todorov, Jeff Boote, and Prasad Calyum about Abilene Measurement Infrastructure (AMI).

04 December 2002 - Eric Boyd presented information on piPEs during Tutorials held at Internet2's Ann Arbor offices.

05 December 2002 - During the All-Staff Retreat, Cheryl Munn-Fremon provided staff with an overview of the working groups that Internet2 supports, how such groups are formed and what the expectations are regarding the output of such groups. Eric Boyd hosted a meeting of software developers, aimed at consensus building on the tools that should be available for use and supported by the Technical Support Group (TSG) in the future.

12 December 2002 - Eric meeting with John Streck and John Moore of NC-ITEC about Bill Wing's work (DOE OR) and continuing it on Abilene with Spirent gear.

04 February 2003 - At Joint Techs in Miami, Eric Boyd gave a presentation on the Abilene Measurement Infrastructure (AMI)/piPEs status; Jeff Boote gave an update report on OWAMP (an implementation of the proposed standard for a one-way active measurement protocol).

05-06 February 2003 - The E2Epi hosted the second E2E Measurement Workshop directly following the Joint Techs