CAIDA ITL Proposal to NSF

Proposal for Development of

University Internet Teaching Labs

Proposed as an Addendum to the IEC Grant

NSF Award Number NCR-9706181

Kimberly C. Claffy and Evi Nemeth, PIs
University of California, San Diego (UCSD)
San Diego Supercomputer Center (SDSC)

http://iec.caida.org/itl/

The Internet Engineering Curriculum (IEC) Repository is a project of the Cooperative Association for Internet Data Analysis (CAIDA). CAIDA is based at the University of California's San Diego Supercomputer Center (UCSD/SDSC). The repository is located at http://iec.caida.org/.

I. Project Summary
II. Table of Contents
III. Project Description
IV. References Cited
V. Bios - available in hard copy proposal
VI. Budget
VII. Current and Pending Support - NSF Form 1239
VIII. Facilities, Equipment and Other Resources - NSF Form 1363
IX. Special Information/Supplementary Documentation - Letter of Commitment from Cisco, Systems

I. Project Summary

In early 1998, efforts began to create an Internet Engineering Curriculum (IEC) repository for high quality network engineering training materials. Support for the IEC is provided by the National Science Foundation (NSF), under NCR-9706181, the Association for Computing Machinery (ACM), and members of the Cooperative Association for Internet Data Analysis (CAIDA).

During early survey efforts, it became apparent that few of the nation's Universities have courses in networking technology and even fewer have facilities for hands-on exposure and training of students on current Internet hardware and software. With donations of equipment from vendors, notably Cisco, and financial support from NSF, we propose to create Internet teaching laboratories at 15-20 of the nation's Universities. This proposal describes our plans to broaden the scope of the IEC to support the establishment of these Internet Teaching Labs (ITL).

The prototype ITL facilities will be made possible by generous donations of networking hardware and training from Cisco Systems (see letter of commitment in Section IX) and network engineering talent to assist in leading this initiative by MCI Worldcom. The IEC Advisory Board will select the schools to be recipients of the lab equipment and training based on criteria outlined in Section IV of this proposal. We ask NSF to provide the "glue" to help put the labs together by providing financial support to cover modest technical, intern and administrative labor; travel; small equipment items; and preliminary training. Additional proposals focusing on more in-depth training and curriculum development may be submitted by IEC collaborators, notably the University of Virginia's School of Engineering and Applied Science.

This proposal should be considered a prototype to a potentially much larger paradigm where major Internet providers together with network equipment manufacturers regularly augment and create University teaching labs by recycling high-speed routing equipment being upgraded by their larger customers. This broadening to more vendors and more participating ISPs can not only add diversity to the labs, but also keep them reasonably current and up to date and provide the commercial sector with a critical flow of undergraduates qualified to move immediately into the Internet workforce.

To kick-off Phase I of this initiative, Cisco is committing over 100 Cisco7000 routers originally valued at more than $10 million. MCI Worldcom is donating the time of a leading engineer from the NSF-supported very-high-performance Backbone Network Service (vBNS). Other vendors and ISPs have expressed a willingness to participate in Phase II of this project and serve as supporters/mentors of select recipient ITL schools.

III. Project Description

1. Background and Motivation

In this section we provide background material on both the need for Internet teaching laboratories in U.S. Universities, and on the Cisco and MCI Worldcom donations that make these initial prototype laboratories possible.

University Networking Curricula

The state of Internet technology teaching at the nation's Universities does not reflect current economic and social realities. Industry demand for labor with Internet expertise far outstrips the supply, and undergraduates who have grown up in the Internet world of the 90's are streaming into computer study programs at our Universities in record numbers. Unfortunately, these students have few Internet technology-related choices in existing curricula. The NSF-funded project, "Internet Engineering Curriculum Repository" (IEC) attempts to help by making it easier for faculty to develop networking courses. The current "Internet Teaching Laboratories" proposal will allow those courses to have a laboratory component with modern equipment.

The motivation for establishing such Internet technology labs stems from several economic and academic factors, all leading to the conclusion that Internet courses and hands-on access to equipment and software is woefully absent in the United States' education system. It is well known that demand exceeds supply in today's explosive Internet marketplace for qualified engineers and associated expertise. The more general high-tech industry reportedly has 350,000 vacancies with increasing pressure on drawing talent from overseas. It is not surprising that college enrollment in computer studies shot up 40% last year, with more students matriculating from high schools with at least some knowledge of, interest in, and experience with the Internet.

The Cisco/MCI Worldcom Donation

In early 1998, MCI Internet Engineering designed an upgrade to most of the routing nodes in the commercial MCI backbone, replacing many of its core routers with newer, faster and higher capacity units. MCI negotiated a contract with Cisco for the new equipment purchase that included a trade-in of existing routers in exchange for the new state-of-the art equipment. These "obsolete" routers have an aggregate original list value of over $10 million, and are comparable to the routers being used in many commercial backbone infrastructures today.

Under the terms of CAIDA's agreement with Cisco, most equipment will be donated directly to the recipient Universities for use exclusively in a teaching/research lab setting. Operational use in the University's networking infrastructure is expressly forbidden since this would compete with Cisco's commercial offerings. Execution of an Acceptable Use agreement by each school, is anticipated prior to the transfer of any equipment.

Each University lab will receive from 3 to 10 routers depending on their need and anticipated usage, see evaluation criteria in Section III 3 below. Cisco is also donating associated software licenses. Some technical support may also be provided by Cisco.

We are also asking Cisco to train, on a space-available basis, faculty or support staff at each selected institution prior to the lab's operation as part of a University class. Cisco is already providing instructors for CAIDA's IEC workshops on routing technology scheduled for August 1999.

2. Description

While Cisco's generous equipment donation makes possible the formation of these labs, simply leaving equipment at the doorstep of Universities will not accomplish the task, and could result in failure if insufficient resources and priority are assigned to the ITL by the recipient schools. A coordinated, funded program is required to work with schools in the planning, design, construction, and initial operation of the Internet labs. Program coordination and engineering services form the two main components of the project. We describe the engineering plan to design, build, implement, and operate the labs, and the program management plan to support the overall activities including inter-lab connectivity and coordination among Universities and associated NSF programs such as CAIDA's IEC.

The construction of the Internet Teaching Labs is described in two phases. Phase I includes the primary elements of the project, during which time awards are made to recipient schools and the Internet labs are constructed with the current set of Cisco-donated equipment. We anticipate future donations of more equipment by industry vendors, leading to diversification and modernization of existing labs and the establishment of ITL facilities at additional Universities during Phase II. We expect the ITL to receive a high degree of publicity and garner attention within the Internet community during Phase I and feel that it is realistic to expect additional vendors to participate in the ITL once it is underway. Nortel, for example, has already expressed interest in participating in the project, by assisting to establish 2-3 high-end University training laboratories focusing on high performance network switching hardware and software.

Engineering Plan

Engineering support is needed for lab design, site review, construction, installation and initial assistance in the operation of the facility. This support is restricted to the ITL lab equipment, and will not include campus infrastructure. The budget for this effort includes funds for a site visit and on-going telephone/electronic technical support for 1 year.

Lab Design

Lab design is a critical component, in part because we see advantages to maintaining a consistent design for all labs. Remote troubleshooting and support will be much easier if the basic lab design is the same at each site.

MCI Worldcom engineers have solid experience in Internet lab design and construction. Kevin Thompson, senior engineer for the vBNS, will lead the design of each Internet lab, with local modifications to the model made according to the parameters of the specific site. Thompson was also instrumental in the establishment of the router laboratory at the University of Virginia (note this laboratory uses Cisco 7000s removed from the previous MCI backbone). Thompson will be supported by Mike Forbes, an engineer working for CAIDA. Forbes is based at the University of Colorado, Boulder where he was responsible for the creation of their Telecommunications (including networking) Laboratory. During the site review, an ITL engineer will tour the candidate lab facilities, determine lab space and environment suitability, and discuss any issues that may arise with University representatives. As we gain experience with the labs, these site reviews prior to design may not be necessary.

Lab Installation

An ITL engineer on-site will supervise the lab installations. Basic equipment installation is expected to take 1 day provided the site is properly prepared. Any special configuration issues will be handled by the ITL engineer in conjunction with the institutions networking faculty contact. Engineers, technicians, professors, and graduate assistants at each awarded school are expected to be fully involved with the ITL team at each phase, including installation and configuration.

Installations will be staggered throughout 1999/2000, beginning with the first lab at the University of California, San Diego. CAIDA will use this laboratory as part of the routing course during the IEC's August Workshop, see http://iec.caida.org/workshop/. The second prototype installations will be at the University of Colorado to support classroom instructure as well as the Winter IEC workshop. Additional installations will be conducted throughout the year following review and approval of a site's application. Our goal is to have all Internet labs under Phase I completed and operational by Spring 1999. Engineering support will be provided for the first year of a facility's operation. Financial support for labs that may be established or augmented under Phase 2 is not included in this proposal.

Program Management Plan

The principal investigators responsible for CAIDA's IEC are Dr. K.C. Claffy of UCSD and Dr. Evi Nemeth, who holds joint appointments at the University of Colorado, Boulder and UCSD. Claffy and Nemeth will maintain technical responsibility for the ITL add-on to the IEC, with CAIDA's Director, Tracie Monk, responsible for administrative management of the initiative.

Day-to-day management of the ITL project will be provided by Kevin Thompson (MCI Worldcom) with support from Mike Forbes, CAIDA. Technical advisory support, including review of candidate ITL recipient Universities, will be provided by the IEC Advisory Board. The Advisory Board members are:

Chase Bailey, Cisco Systems
Steve Bellovin, AT&T Research
Scott Bradner, Harvard University
Randy Bush, Verio Networks
John Connolly, University of Kentucky
Jon Crowcroft, University College London
Jim Kurose, University of Massachusetts
Rick Wilder, MCI Worldcom

The modest support requested from NSF will be used to facilitate the dissemination of information to U.S. Universities concerning the ITL initiative and to assist in the establishment and initial operation of 15-20 facilities.

Staffing

Day-to-day leadership of the ITL effort will be provided by Kevin Thompson, a senior engineer whose time is being graciously donated by MCI Worldcom. Thompson will oversee the construction of the labs including providing overall program management and coordination, managing lab construction, University interaction, schedules, general implementation issues, and coordination among Universities and ITL. Thompson will be assisted by a CAIDA engineer whose time will be funded through NSF's support of this project. This engineer will be responsible for facilitating lab design, site review, construction, installation and operation.

Funds are also requested to provide stipend support for undergraduate and graduate interns at each sites to assist with installations and/or be responsible for operation of the teaching lab. The ITL engineer will coordinate with on-site faculty responsible for supervising the student interns.

Additional funds will cover administrative (managerial, clerical, systems) backstopping of this effort, and the participation of one or more dedicated undergraduate interns based at SDSC.

Equipment

Laptop computers will be used to support the fieldwork of both engineers and the coordinator. Desktop PCs will be used as support systems for management and administrative duties. The existing CAIDA web server will be used.

IEC Coordination

The importance of close coordination with the IEC cannot be overstated. The general goals of IEC and ITL are the same, namely, the proliferation of Internet technology classes at the nation's Universities and the emergence of the laboratory infrastructure to support them.

The IEC repository, in addition to University classes, workshops, and tutorial materials will have a special section on laboratory exercises appropriate for networking classes using ITL or other laboratory facilities.

The ITL will be fully integrated into IEC activities, especially the IEC workshops, where ITL will provide lab equipment for use during the workshop sessions. ITL schools will be encouraged to participate in IEC activities, including but not limited to: attending workshops; contributing material to the IEC repository where appropriate; and working to integrate IEC studies and materials into the school's Internet lab courses.

Potential distance learning and collaborative lab exercises among schools should arise from these efforts. An example is Dr. Nemeth's course on Network Systems that was taught simultaneously at the University of California, San Diego (UCSD) and at the University of Colorado at Boulder in the fall 1998 semester using the vBNS network as the transport medium. The lectures were multicast across the vBNS originating at UCSD for the first half of the semester and from UC Boulder during the second half. The class involves lab exercises but because the equipment only existed at the UC Boulder site, labs could not be required. The existence of ITL labs will provide a common base for future distance learning experiments between participating schools.

Information Dissemination

An ITL web site will be established in connection with the existing CAIDA and IEC sites. The purpose of this web site will be to publicize and disseminate information on ITL activities (past, current, planned). Information will include application forms for participation (either as a recipient school or as a vendor), announcements of schools selected as ITL recipients, status of laboratories being established at the respective schools, and details on operational labs, including course descriptions and references to any lab work and materials. The web site will also serve a tracking function for all equipment under the auspices of IEC. Complete lab specifications will be available on the web site. This information can be used by schools who were not awarded ITL labs, but who may have other funding to construct an Internet lab. Lab specifications will also aid IEC course developers in designing future lab courses.

Schedule

Construction of the first two laboratories (UCSD, and UC-Boulder) will commence immediately upon award notification from NSF, as will dissemination of the selection criteria for other ITL recipients. Construction of additional ITL facilities should commence within the second quarter of this effort, with the goal of preparing labs for use in the Winter and Summer 2000 semesters.

Additional NSF Support

NSF may wish to consider funding 1-2 stipends for graduate assistants at each ITL school to support and maintain the lab operations, assist student lab exercises, and develop new exercises for future lab courses. Support for these students would be requested by the individual Universities directly from NSF. However, close collaboration with the IEC project team and other ITL schools would be encouraged.

Discussions are also underway with the first ITL-type school, University of Virginia, about their plans to develop a Lab Manual on use the ITL equipment and related resources. UVA, under the leadership of Jorg Liebeherr, is also considering submitting a proposal to NSF to facilitate this effort and to support a seminar training instructors at UVA and other ITL institutions on the use of the Lab Manual. This proposal would be from UVA to NSF (without UCSD's financial involvement), but should be viewed as an extension of the IEC project scope. CAIDA's IEC staff are collaborating with UVA and feel that their efforts are consistent with the goals of the ITL.

3. Participating Universities

In this section we describe three proposed prototype laboratories and the selection criteria that will be used to choose the additional schools to receive the Internet Teaching Laboratories.

Lab Prototypes

As a precursor to this proposal, MCI and Cisco arranged with the University of Virginia to create a prototype Internet lab at UVA's School of Engineering and Applied Science. UVA provided space in a pre-existing computer room with raised flooring and proper environment and power to house the lab equipment. UVA assigned Prof. Jorg Liebeherr to operate and maintain the lab. Prof. Liebeherr has an existing lecture-based course on Internet technologies that he began teaching last year, and he plans to integrate lab exercises into this course. MCI Worldcom and Cisco engineers installed the equipment at UVA, configured and debugged all routers and connected devices prior to its official inaugeration in April 1999.

Labs will be built at UCSD and UC Boulder to support both the IEC workshops and University networking classes. Since all ITL facilities will have the same basic design, faculty from ITL schools will use the same equipment at the IEC workshops as they have at home. Technology transfer should be easier and more complete.

Qualification Criteria for Labs

The selection of the schools to participate in this program will be based on a number of criteria. To participate, schools are required to integrate the lab into one or more specific undergraduate and graduate networking classes. Schools may also want to leverage the lab capabilities in ongoing and planned graduate research opportunities. This is a fine application of the ITL labs, however the primary function should remain as teaching laboratories to maximize exposure of the equipment to undergraduate engineering students.

We would like to target two types of schools:

big research schools with faculty in networking;
smaller teaching institutions with faculty and Departmental support

The top research institutions often lack laboratory facilities such as the ITL will provide or they lack the funds/inclination to build them. These institutions already have networking courses and faculty expertise. Locating labs here will have a high probability of success.

But, smaller teaching institutions are often more appreciative of help with laboratory equipment and their energetic faculty are more willing to spend time on curriculum revision. Locating labs in these smaller colleges and Universities will have a strong, positive impact on existing curricula as well as good chances for success.

Schools will be chosen to receive Internet Teaching Laboratories by the IEC Advisory Board. The proposed criteria for evaluating whether a University should be a recipient of an ITL may include:

identification of one or more one keenly interested faculty members who will agree to implement curricula utilizing the facility;
the existence of undergraduate courses where an ITL laboratory component can be added or expanded to use the lab in the near future;
graduate courses or research projects where an ITL laboratory is needed;
commitment of the Department to provide support for faculty involved with the labs (release time for curriculum development, student staff to help with laboratory exercises, etc.);
commitment of the University to provide adequate space, power, cooling facilities to house the equipment in a healthy environment;
commitment of the University to assign staff (career or student) to be responsible for local maintenance and operation of the lab;
commitment from the University to a plan to encourage cooperation between competing Departments (Computer Science, Electrical Engineering, Computer Engineering, Telecommunications) in using and scheduling the labs
identification of any plans to make the ITL facilities available to other community groups, e.g., other colleges and Universities, junior colleges or even high schools in the area.
identification of additional resources (if any) that will be available for the ITL, e.g., racks, console terminals, other networking equipment;
identification of any further contributions (network engineering curricula, training, manuals, outreach, etc.) that the applicant University is willing to make to the IEC project or to participating ITL schools, e.g., serving as a mentor for newly formed lab facilities.

An application form covering these criteria will be posted on the ITL homepage at iec.caida.org/itl and will be distributed by mail to approximately 150 Computer Science Departments throughout the U.S.

4. Potential Impact

We summarize the potential impact of Phase I of the Internet Teaching Laboratories project below and look ahead to issues that fall into Phase II.

Summary, Phase I

The long-term goals of strengthening and growing Internet Teaching Labs are shared by the IEC Project and its collaborators, notably MCI Worldcom and Cisco. The potential impact of ITL facilities at our top Computer Science Departments across the nation is huge.

The eventual degree of integration and absorption of the labs into the schools' engineering curriculums will determine the measure of success for ITL. The question may center on the lab's effect on the University engineering curriculum and the ability of hand-on ITL exposure to enhance the technical grounding of graduates. A ng education in any engineering discipline invariably includes significant lab work, and Internet engineering should not be an exception. The IEC's goal of dynamic, commercially-relevant Internet engineering curriculum components is a challenging and invigorating goal that we share. Such penetration and influence at the country's best engineering schools is the long-term objective of ITL and ultimately, the grail upon which the project's success should be judged.

On a more pragmatic level, we plan to request that participating institutions participate in lab evaluations, distributing forms to students interacting with ITL labs. These evaluations will be separate from regular course evaluations administered by the Universities and will ask the student to evaluate the lab's effectiveness. These evaluation forms will be processed and summarized by CAIDA/IEC staff, and the analyses added to the web site regularly. Faculty and teaching assistants will also be asked to submit evaluation forms; selected instructors will be interviewed for more in-depth feedback and criticisms.

Futures, Phase II

The initial Internet Teaching Laboratories are expected to be Cisco and MCI centric due to the critical equipment and manpower contributions these institutions are making during Phase I. We are confident that by the conclusion of Phase I, the benefits associated with enhancing the emerging Internet engineering workforce and cooperating with U.S. Universities through ITL and related endeavors will be clear. Given the degree of interest being expressed by other vendors and ISPs today, we envision that this ITL component of the IEC will be self-supporting by Phase 2 and that several additional organizations will be contributing resources.

Several other countries have expressed interest in the IEC repository and IEC workshops. We anticipate that this will also occur for the ITL facilities as well, increasing the Internet engineering knowledge and training opportunities world wide.

5. Results from Prior Work

This project is an add-on to existing efforts of the Internet Engineering Curriculum Repository and CAIDA. For information about the progress of these two efforts, see the IEC's Year 1 and Year 2 status reports and CAIDA's 1998 Annual Report.

III. References

The activity proposed in this project is training and outreach related, with a significant focus on building university capabilities at the undergraduate level. Descriptions of current conditions facing U.S. universities can be found in the Administration's Next Generation Internet (NGI) Initiative, the Information Technology for the Twenty-First Century (IT2) Initiative report, and the President's Information Technology Advisory Committee (PITAC) report. Additional documentation is also available in numerous NSF publications relating to conditions in higher education.

IV. Bios

Bios for the co-PIs, Dr. K.C. Claffy and Dr. Evi Nemeth, are being submitted under separate cover.

VI. Budget

Budget and financial materials supporting the Internet Teaching Lab add-on to the IEC project is attached. The project would run for 14 months, from August 1, 1999 thru September 30, 2000.

Staff support requested under this effort includes a part-time Internet engineer, 2 half-time student employees to help with installations and integration with the broader IEC project, plus administrative/management support.

Travel will consist of site visits to each of the recipient ITL institutions for site review and/or installation.

Funds are also requested to support two small workshops ($10,000 total) with stipends of approximately $1,000 each to assist in defraying participant travel expenses. These CAIDA/IEC sponsored workshops will focus exclusively on routing and routing exercises for undergraduate and graduate courses in networking using ITL facilities. We are budgeting for 20 inexpensive, free-standing equipment racks at $200 each to house the equipment at each site. Additional funds (roughly $7,500) are requested to cover fiber, cables, and miscellaneous equipment. Each site will be expected to supply dumb terminals to serve as consoles for the routers.

Approximately 10 institutions are expected to receive graduate student stipends for teaching assistants to help with laboratory components of the networking courses during the first year. CAIDA envisions providing some financial support for students at select institutions from its membership support funds. Additional requests for support of undergraduates and graduates working on the ITLs may be made directly to NSF by participating institutions.

1030 Forms are being submitted in hard copy.

VII. Current and Pending Support

-NSF Form 1239

VIII. Facilities, Equipment and Other Resources

- NSF For m 1363

IX. Special Information/Supplementary Documentation

- Letter of Commitment from Cisco, Systems

Last updated 7 June 1999

Proposal for Development of University Internet Teaching Labs