Written by Dharma Akmon

Written by Dharma Akmon August 31, 2004

Backbone Network Infrastructure (BNI)/Abilene Record Groups

The BNI division of Internet2 is responsible for developing and testing advanced network services that support the needs of its members. Abilene, the backbone network of Internet2, is the most prominent project of this area, but FiberCo and the Quilt are among others important to the goals of Internet2. While many of the divisions of Internet2 carry out their work through working groups, the core activity of BNI is conducted either by Internet2 staff or through participation in other projects and initiatives. For the latter reason, it is sometimes difficult to draw the line between the work of Internet2 and that of other organizations. However, with respect to BNI and Abilene, Internet2 sees its role as that of a policy maker rather than as the organization carrying out the day-to-day work of running a network. For that reason, I have tried to focus on those activities that reflect Internet2 as an enabler.

Working Groups- BNI currently has only one working group, IP Optical, and it is has its own website at The website for the working group includes the charter, past meetings, presentations, a list of current members and slots for meeting minutes. This group also has a listserv that is archived.

Presentations- Some presentations are linked through staff members’ bios, but generally all can be found on individuals’ computers. Key people in BNI are Guy Almes, Steve Corbato, Matt Zekauskas, and Rick Summerhill. Presentations will also be found on relevant meeting pages as all Internet2 meetings are archived online.

Internet2 Technology Evaluation Centers (Itechs)- There are four evaluation centers that serve as national network testing laboratories. Located at Ohio State University, North Carolina State University, San Diego Supercomputing Center, and Texas A&M, each has a report of activities online at

BNI Projects- BNI is engaged in a number of projects, some of which are run through Internet2 and others of which are led by outside groups. The Internet2-led projects each have a website, which contains most of the records available for a project (aside from email communication). The projects are:

• FiberCo- This effort is dedicated to supporting regional fiber optical networking. The website contains press releases, news clips, Level3 Network maps, frequently asked questions and recent presentations.
• Abilene Observatory- The Abilene Observatory supports the collection and dissemination of data about the Abilene network. There are a number of research projects that use the Abilene Observatory. Descriptions of the observatory and related research projects are available at however, the project and data collection are performed by other institutions and are therefore stored on outside websites.
• Manhattan Landing Exchange Point- MAN LAN is a high performance network exchange point that facilitates peering among research and education networks. A synopsis of the project can be found at
• K-20 Initiative- This initiative was formed to make the Abilene network available to a broad educational community that includes K-12 schools and community colleges. Meeting documents, newsletters, press releases, and a list of people involved can be found at
• HOPI- The Hybrid Optical and Packet Infrastructure project seeks to use hybrid packet switching and optical lambdas to scale up the network up for the future. The website for the project at includes a summary of the project, white papers, and updates. While currently many of the documents for the project are on a password-protected portion of the website, these will soon be available to the public.

Abilene- Abilene is the backbone network of Internet2. Indiana University is responsible for the day-to-day running of the network, while Internet2 assumes a more prominent role in developing policies.

• TAC-In 1998, Abilene formed a technical advisory council made up of important CIOs to provide advice on the project as well as act as a sounding board for different ideas the staff has about the direction of the network. The Abilene TAC has a listserv that is automatically archived, and this is the primary repository of all records related to this group. Minutes from meetings as well as documents created by the TAC are also kept in the listserv archive as attachments to emails.

Abilene Planning Committee- The Abilene Planning Committee was formed in 1998 to plan how the network would develop and operate.

• Emails- this planning committee has a listserv that automatically archives all emails. Matt Zekauskas mentioned that it might contain some sensitive information related to companies that the project is involved with.
• Meeting Minutes- Linda Charlesworth keeps the official minutes from the planning committee from 2001 to the present. The person previously responsible is no longer at Internet2 and the minutes left with her. My impression is that the minutes are spotty since a committee member reported that the group is no longer rigorous about taking minutes. However, Matt Zekauskas tries to take almost word-for-word transcripts of meetings he attends, including the Abilene Planning Committee meetings. Although they are not official, he said that they are a very good record of what happens at meetings.

Contracts and Agreements- Abilene holds all of the contracts for members that connect to Abilene. These are kept only in paper form and date to 1999.

Archived Planning Files- Because there have been no strict recordkeeping policies, early documentation for Abilene will be difficult to assemble. Heather Bruning has kept a very small paper file about the planning of first deployment of Abilene. Guy Almes is the most likely person to have early materials, but we have been unable to meet. Some documents exist in the old eRoom, please consult Excel file for more detail.

• Network maps- Heather Bruning is responsible for keeping network maps, which are updated on a regular basis and show the major connecting points across the country for Abilene. Network maps will also be found in various presentations given over the years.

• Rack Engineering documents- The rack engineering documents show how the original engineering of the network was conceived. Matt Zekauskas reported that many staff would have these, and they can also be found in the old eRoom.

• NOC setup- Indiana University is the Network Operation Center of Abilene and Zekauskas stated that documents of the selection of the NOC “might exist somewhere.” Some paper records were found in the office of Heater Bruning, but Guy Almes may have more substantive documentation (I have not been able to meet with him).

• QoS-Quality of Service was a project initiated to support the development and deployment of advanced network applications through the use of IP traffic differentiation. Although the project was disbanded, it was seen as an important initiative and has been one of the few projects at Internet2 that has failed. This group has an archived listserv as well as a working group website:

• Joint techs meetings from 1998/99- Matt Zekauskas said in an interview that meetings from this period might have presentations about Abilene in the early planning and implementation stages. However, since these meeting documents have not been well maintained online, staff presentations will probably yield the best documentation.

• Deployment Focus Group- This group was formed to procure the necessary equipment for deployment in the Abilene backbone network. They had a listserv discussion group that should be archived.

• Development of Conditions of Use- Documents that trace the development of the conditions of use for Abilene can be found in the old eRoom. Please see Excel sheet for more detail.

• Second Generation Abilene- The second generation of Abilene was deployed in early 2002. There are Word documents that describe tests run as well as summaries of their results. Some of these documents are very sensitive because there was some controversy over who would provide the routers.

Summary

The records for BNI are generally scattered amongst individuals working in this area. Although important records can be found online, I also discovered that there are many gaps in documentation, especially in the early development of Abilene (1998). Staff reported that much has likely been lost from this time, but presentations, Internet2 member meetings, and the archived listservs will all provide valuable information on how BNI and Abilene developed. The emails of the Technical Advisory Council in particular will be of high research value because they likely show how Internet2’s members (CIOs from universities) thought the network should be implemented. More technical documents, such as the rack engineering plans and network maps, which have been well-maintained by Internet2, will be of interest to those who understand network engineering.

One of the most difficult aspects of archiving this area will be deciding where to draw the line, since so much of the area’s work is conducted via its involvement with outside groups. This also points to a significant issue in the organization as a whole since Internet2 sees itself as the organization that helps research institutions make the internet better. So, while it might be very interesting to archive, for example, the network activity of Abilene, these records are kept by the NOC (Indiana University) and represent work that falls outside of Internet2’s primary role.