telecom@eecs.nwu.edu (TELECOM Moderator) (02/13/91)
[Moderator's Note: The attached was sent along by Jody Kravitz and is being sent out as a special mailing because of its size. Regarding the flow of stuff into telecom, readers should be advised I still have over 100 messages backlogged in the queue of things to print. Most of it will get out in the next two or three days .... all will be redated as needed to keep it from expiring in comp.dcom.telecom. Again please: DO NOT send further messages to the group until toward the weekend. Thanks. PAT From: foxtail!kravitz@ucsd.edu Subject: Use of T3 for Packet Switched Networks on the Increase Date: 13 Feb 91 7:18:12 GMT Organization: The Foxtail Group, San Diego, CA This month's California Education and Research Federation Network (CERFnet) newsletter talks about the installation of T3 circuits in NSFNET. NSFNET and CERFnet are part of the packet-switched "Internet". The growth in use of T3, which operates at 45Mbps, may be of interest to some of the readers of the TELECOM Digest. I've excerpted the relevant articles from the newsletter and included them below: CERFnet News February 1991 Volume 3, Number 1 INSIDE THIS ISSUE: * The new faster NSFNET reaches CERFnet SDSC installs new 45 Mbps connection to the NSFNET. This is part of the NSFNET migration to 45 Mbps. This article also discusses the benefits to CERFnet users. .... articles deleted .... * Initial T3 deployment in place on NSFNET This article discusses NSFNET's new 45 Mbps (T3) backbone and future improvements of the T3 architecture. .... articles deleted .... Staff for this issue of CERFnet News includes: Editor Advisors Karen McKelvey Susan Estrada Robert Morgan Writers Contributors ESnet staff Mike Beach Ken Horning Rachel Chrisman Paul Love Carlos Robles Cathy Wittbrodt CERFnet News is published monthly by the California Education and Research Federation Network (CERFnet). CERFnet is a mid-level network linking academic, government, and industrial research facilities throughout California. CERFnet receives partial funding from the National Science Foundation (NSF), operating under grant number NCR8819851. Any opinions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the views of NSF, other funders, General Atomics, SDSC, or CERFnet. If you would like to receive CERFnet News or would like further information about CERFnet, please send your request to help@cerf.net, or telephone 800-876-CERF, or contact the CERFnet office at the following address: CERFnet c/o San Diego Supercomputer Center P. O. Box 85608 San Diego, CA 92186-9784 A NEW FASTER NSFNET REACHES CERFNET SDSC installs new 45 Mbps connection to the NSFNET. by Paul Love During the year-end holidays, the San Diego Supercomputer Center (SDSC) installed its new 45 Mbps (T3) connection to NSFNET. This was accomplished during a visit to SDSC by Hans-Werner Braun, then the Merit Principle Investigator on the NSFNET project. (Braun has since become one of SDSC's networking specialists.) CERFnet users can now enjoy faster service and gain access to even greater computational capabilities. The NSFNET T3 circuit at SDSC uses a fiber that was recently installed from UC San Diego and SDSC to the local MCI office. This fiber connection bypasses the local telephone office which reduces rates for connections to MCI. The fiber installation was possible because most of it runs through land owned by UC San Diego, Chevron, and General Atomics, who cooperated in its installation. The fiber also means new technologies can be added quickly. For example, the CASA testbed network (see article in CERFnet News, Aug-Sept 1990) of the National Research and Education Network (NREN) will use connections running at 1 Gigabit. This is only available over fiber. SDSC is a node on the CASA network. As of January 25, traffic on the T3 from SDSC was limited to Merit (in Ann Arbor, MI) and the National Center for Supercomputing Applications (in Urbana-Champaign, IL). Today, connections include Stanford University (Palo Alto, CA) and the Pittsburgh Supercomputing Center. Initially, six of the thirteen NSFNET sites were scheduled to receive T3 service in addition to their T1 service. Also, two new sites (Cambridge, MA and Argonne, IL) were scheduled to receive only T3 service. At the FARNET meeting in January, Steve Wolff of NSF reported that the agreement with Merit/MCI/IBM has been modified. All sixteen sites on the NSFNET backbone will receive T3 service. These installations are expected to be completed by the end of the year. The T1 network will be dismantled when all of the sites are operating T3 service. Benefits for CERFnet users The T3 midgration means that CERFnet users will not experience any degradation of service during 1991. It was projected that some of the links on the T1 network would become congested this year if additional bandwidth was not provided. Also the new T3 network will make new services available. Remotely mounted file systems will seem much more like those mounted just across your local Ethernet. Distributed computing across the country will now be practical. The new network will make document and data retrieval a faster, simpler, and common operation -- a necessary service for users as more full -- text libraries and databases become available via the Internet. Effect on mid-level/local networks The growth of most mid-level and local networks is another important facet of the NSFNET migration to T3. Most mid-level networks have a high percentage of T1 links. The new T3 connections will keep a mid-level's connection to the NSFNET backbone from becoming a bottleneck. Also, the T3 connections will provide better service to local network resources, especially where the supporting LANs use FDDI. Summary Merit has been tracking the growth in network usage for the last several years and offers these eye-opening statistics. In November 1988, the network carried less than 400 million packets. In November 1989, the rate was 2.1 billion. And, by November 1990, the rate was over 3.8 billion. While the percentage increase is falling, the absolute number of packets carried has grown each year by the same amount: 1.7 billion. If this growth continues, we can expect NSFNET to carry over 6.5 billion packets by November 1991. The NSFNET migration to T3 will keep the backbone ahead of network usage patterns. It will allow users to use new network-intensive services as they become available resulting in a positive impact on scientific inquiry and industrial R&D. * INITIAL T3 DEPLOYMENT IN PLACE ON THE NSFNET by Ken Horning [Editor's Note: This article is reprinted from LINK LETTER, V3 N 5, December 1990. Ken Horning works for Merit/NSFNET.] Operational deployment of NSFNET's new T3 backbone was started in the final months of 1990. T3 installations are now complete and ready for operational traffic at the backbone end nodes in Ann Arbor, MI, Urbana-Champaign, IL, San Diego, CA, and Palo Alto, CA. "This upgrade again respects the National Science Foundation's (NSF) commitment to keep NSFNET the world's leading computer network for the support of research and education," said Dr. Stephen S. Wolff, Division Director, Division of Networking and Communications Research and Infrastructure, at the NSF. "New applications that were not feasible on slower networks will be possible with the availability of T3 bandwidth." Production ready Prior to their installation, the T3 connections at the four installed nodes were thoroughly tested. Testing procedures included continued verification of hardware, software, and circuits to evaluate reliability. A suite of testing tools and procedures has also been created which will facilitate the installation of the T3 connections at the remaining nodes. The model developed for high-speed backbone transmission involves a new generation of Nodal Switching Subsystem technology developed by IBM. Advanced circuit technology for the T3 upgrade is being provided by MCI. Future improvements The architecture for the T3 network is utilizing a collection of IBM Core Nodal Switching Subsystems (C-NSS) within the MCI infrastructure, forming a cloud of co-located packet switching capability. Exterior Nodal Switching Subsystems (E-NSS) are located at client sites and connect into the C-NSS cloud. With the deployment of the new T3 architecture, the node packet switching performance will improve significantly. The initial T3 deployment employs an Ethernet interface to the local area network, providing material performance improvement compared to T1 NSS performance. As the NSFNET partnership completes the FDDI interoperability testing and deploys FDDI with the new technology, even more signi^cant performance improvement will be realized. Additional new technology due in '91 Later in 1991, the partnership plans to deploy new technology which will use intelligent subsystems for the extended interfaces. These subsystems or powerful RISC-based adapters utilize bus master and slave capabilities on high bandwidth implementations of the microchannel to achieve very high-speed card-to-card forwarding with no system intervention. Coupled with optimized distributed protocol code, these systems can achieve very high throughput rates. IBM's RISC on RISC architecture utilizes RS/6000 RISC chipsets for the control processor and a 25 MHz superscalar, RISC embedded controller with on-chip cache and data RAM for the adapter engines. The new technology with on card packet forwarding will dramatically improve the performance on the T3 network. * CERFNET NEWS AVAILABLE VIA ANONYMOUS FTP Issues are available via anonymous ftp to NIC.CERF.NET in the subdirectory cerfnet/cerfnet_news. *