Connectivity Providers Database

Date: Tue, 31 Oct 2000 08:52:28 +0200 (SAST) From: Antoine Bagula <bagula@cs.sun.ac.za> To: sgoldste@nsf.gov Cc: joelja@nsrc.org, klensin@nsrc.org, Brian Candler <B.Candler@pobox.com>, Randy Bush <randy@nsrc.org>, Steven Huter <sghuter@nsrc.org>, mbuyimuk@sciences.unikin.net, mbungu@sciences.unikin.net, hugi@oregon.uoregon.edu, js@rcp.net.pe, bombambo@sciences.unikin.net, didier@unikin.net, zita@nsrc.org Subject: Academic Network of the Democratic Republic of the Congo Message-ID: <Pine.LNX.4.10.10010310840240.32200-100000@chivaz.cs.sun.ac.za> Dear Dr. Goldstein, We met some years ago during Geneva's INET conference. At that moment, we were still seeking how to set up the academic network of the Democratic Republic of the Congo, beginning with the university of Kinshasa. I am presently working towards my PHD at the university of Stellenbosch in South Africa. The reason why I send you this message is to inform you and thank you for the contribution of the NSF through the NSRC to the set-up of the academic network in Congo. Please find hereafter a brief story and the state of the academic network of the Democratic Republic of the Congo. Best regards Antoine Bagula. ***************************************************************************** I. The Congo Scientific Network (CS-NET) and the University of Kinshasa's Connectivity to the Internet A. Why the project was initiated When we started this effort, universities in the Democratic Republic of the Congo had no Internet access. Congolese scientists were not able to participate in the rapidly growing use of the Internet for collaborative scientific research, and communicating with colleagues doing similar work. The first TCP/IP link to Kinshasa in the Democratic Republic of the Congo was established in October 1996. The initiative came from a commercial European company which installed a 64kbps satellite link from Kinshasa to UUNET/Belgium. However, the prices applied by the local service provider were so high that only a few foreign companies, ex-patriates, and embassies could afford to pay for Internet access. The price in March 1998 for an unlimited PPP dial-up account was $3,540.00 per month. That is not a typo. Requests from Unikin for an educational discount were not accepted at this time. The situation was hopeless for the Congolese scientific community for which this new "Privilege Internet" was a dream. Congolese researchers and students had to find alternate means of communication in a period when public communication facilities (telephones, faxes, telex) in almost all the educational and other public institutions were so degraded that everyone used commercial fax and telephone operators. At that time, a fax communication cost $5 US per page for sending and $2 US per page for the reception. An international telephone call was $3 US per minute and an email message cost $1 US to $2 US per page for sending and reception, if you could find someone who would provide the service for you. Considering the salary of a professor in Kinsasa (about $80/month), these tariffs were so high that the scientific community sadly had to accept our frustrating isolation from the rest of the world's scientific community. B. Beginnings of collaboration with the Network Startup Resource Center via Jose Soriano at an International Centre for Theoretical Physics meeting in Trieste, Italy The Congo Scientific Network project was launched as a non-commercial initiative to overcome this isolation. It was inspired from other projects and work done during participation in workshops organised by the radio communication unit of the Abdus Salam International Centre for Theoretical Physics in Trieste, Italy. Its primary objective was to provide the scientific community in Kinshasa with minimal Internet connectivity (email) at affordable prices. Meeting with Jose Soriano of the Red Scientifica Peruana at the ICTP proved to be invaluable to the effort. From the experience of Red Scientifica Peruana, which had been a recipient of technical assistance from the NSRC back in the early 1990s when establishing Peru's first connection to the Internet, we began to flesh out the possiblities. Soriano described Peru's experiences with developing a basic, store-and forward email system using the Unix-to-Unix CoPy programs to transfer files and mail between UNIX systems. He then introduced me to Randy Bush and Steve Huter of the NSRC and we requested assistance to establish a UUCP connection over dial-up telephone lines to the Congolese Scientific Community. C. Overcoming obstacles to design and build a system The earliest communications with the NSRC to plan the execution of the project were by email from Kinshasa using a couple of different email accounts, depending on the availability and willingness of friends to help for free, or by paying $1 to 2 US per page to local email users who were using a European ISP. Replies from Steve or Randy followed the same friend's paths who called me or used other friends for delivering messages to me. It sometimes took two to three weeks to deliver an urgent message, and obtain the reply. During that period, Steve enlisted the help of Brian Candler, who was also a former recipient of NSRC technical assistance while establishing the first electronic mail system in Belize in the early 1990s. (The NSRC had provided a UUCP email feed over dial-up lines to the University College of Belize as part of the Caribbean University Network project in the early and mid 1990s.) We needed technical information, so to get started, we obtained some networking textbooks, and technical documents. The NSRC sent to Kinshasa a large shipment of books in both English and French that were donated to them by O'Reilly and Associates, and detailed notes on configuring a UUCP-based system that were written by Brian Candler. The shipping costs for the books were paid by a small grant to the NSRC from the National Science Foundation. We installed Redhat Linux on a machine in Kinshasa and configured UUCP with assistance from Steve and Brian using our network of family and friends to send and receive electronic messages, as needed. D. The Congo Scientific Network Since it was expensive and difficult to register domain names using our ccTLD (.cd), I asked Steve and Randy to help us register two domain names that we selected - unikin.net, and cs-net.org. The NSRC provides nameservice for these domains, and hosts a web site for www.cs-net.org, which we use to encourage collaboration with other Congolese researchers around the world. One of its objectives is to make information available about scientific research, and training projects, happening in Congolese universities and research centers. The first volunteers who are sharing information through this effort are doctoral students who are helping colleagues in Kinshasa benefit from their overseas research experiences by sharing their work with scientists and students in Kinshasa. The collaboration aims also at helping Congolese scientists who are overseas to keep contact with their mother country and prepare them for working in the Congo upon completion of their Ph.D. programs. Three working groups have been established so far. 1. A networking technologies work group 2. A water research work group 3. An applied mathematics work group II. Meeting at INET '98 in Geneva Meeting in person in July 1998 at the INET Network Training Workshop in Geneva, Switzerland was the first opportunity for us to work together in person. Brian and Steve were staff members of the workshop and I was a student in Track 1. It was not part of the main curriculum in the workshop, but we did prototype configurations for UUCP over dial-up telephone lines during evening sessions in Geneva. The NSRC donated a US Robotics Courier modem to Unikin for the international connection, and two US Robotics Sportsers for use at the University of Kinshasa. III. Testing of dial-up connection upon return to Kinshasa Tests for the dial-up connection started in August 1998. Kinshasa's server attempted to call the NSRC's machine in Oregon (uucp.nsrc.org). The first barrier we hit was the local telephone infrastructure. We were unable to obtain an international telephone line to dial out from the office where our server was located. So for testing purposes, the NSRC initiated the calls from their server at the University of Oregon. It is typically three to four times cheaper to originate the calls in the USA, so it's much more cost effective to arrange an automated polling schedule and operate the system this way. But as we did not have an easy way to pay the monthly phone bills, we attempted to do manual polling from Kinshasa in the beginning. A. Brief overview of telecommunications infrastructure in Congo Kinshasa is a city of about 5 million inhabitants with almost all of the working cabled telephone lines concentrated in an area of about 25 square kilometers around the central post office. Kinshasa is currently experiencing the splendor of cellular phones and, as is typical in developing countries, the old reliable cabled infrastructure is being neglected in favor of the cellular phones. Unikin, the first university to be connected to the local infrastructure, is about 25 kms from the central post office. It has a microwave link, but its power infrastructure had not yet been adapted to the growth of its population. Unikin has about 20,000 students and 500 teachers for a power infrastructure that can serve only half of this population. Finding a direct telephone line from Kinshasa at an affordable price was problematic. To dial out internationally from Kinshasa, one must use prepaid cards, or a cellular phone, or a satellite phone. The prepaid card system is somewhat difficult to configure for international dial-up via modem. Modifications in system files have to be made for each new card used and some problems occur for the modem's time synchronization to the card's operator voice replies. The cellular and satellite phone alternatives were so expensive that they were not options worth pursuing. B. Other non-technical barriers that inhibited progress Right after we successfully tested the UUCP connection dialing from Oregon to Kinshasa, political turmoil in Kinshasa broke out and interfered with our ability to continue establishing the network until peace was restored. We were locked out of the office where our server was located, and it was not safe to try to go there. Then, just as the local political situation in Kinshasa stablized, there was a three week power outage in Kinshasa! Shortly after the daily life in Kinshasa stablized, there was a reduction in pricing for PPP dial-up service with our country's sole ISP. We decided to get a PPP account and do UUCP over TCP/IP, which would ultimately be cheaper, and avoid some of the international dial-up hassles. IV. Working together at INET '99 in San Jose, California A. A lot of unproductive time had passed due to the political instability during the previous months. Nearly a year had passed, and it was time for the INET '99 Network Training Workshop in San Jose, California. I and another aspiring network operator from the Unikin team, Didier Rukeratabaro, attended the workshop to enhance our technical skills to help build our networking capacity. During evening sessions, we worked with Brian and Steve again, and practiced configuring UUCP over TCP/IP. We were able to test our configuration with the uucp.nsrc.org machine housed at the University of Oregon Computing Center from the INET Track 1 classroom in San Jose. B. Establishing a working system Upon return to Africa, I went to South Africa to complete my doctoral studies, and Didier took over maintenance of the Unikin networking efforts. Didier set up a local mail server on a Linux box to distribute email to individual users at the Faculty of Sciences, Faculty of Economics and others, doing UUCP over TCP/IP. In response to a request from the Faculty of Sciences, the NSRC sent Unikin a large shipment of computing and networking equipment, and more books for the the small technical library housed at the Faculty of Sciences. Pentium PCs for servers, hubs, a switch, ethernet cards, ethernet cabling, modems, a multiport serial card, and memory and other spare parts to spruce some of our old PCs were donated by the NSRC. This equipment has been invaluable in enhancing our system. There are now about 500 users of the Unikin network. Four subdomains have been established for various departments at Unikin that are using the system. sciences.unikin.net for the Faculty of Scienecs economie.unikin.net for the Faculty of Economy polytech.unikin.net for the Engineering department esp.unikin.net for the Ecole de Sante Publique (public health) C. University of Kinshasa's network topology +---------------+ |Global Internet| +---------------+ | | +---------------+ | UUNet Belgium | | 194.7.2.82 | +---------------+ | | Asymmetric satellite IP link | (64kbps/256kbps) | +------------------+ | Inter-Connect | .129 194.7.233.128/25 |(ISP in Kinshasa) |-------------------------------| | 194.7.233.33 | | +------------------+ | | | | | .195 DHCP | .33 | | | | UUCP Client for Unikin | 194.7.233.32/27 over TCP/IP |--------------------------- | | sciences.unikin.net 194.7.233.44 | (unikin.unikin.net economie.unikin.net wireless IP link) | | | | economie.unikin.net UUCP mail feed via uucp.nsrc.org, | serving unikin.net and cs-net.org esp.unikin.net | | | polytech.unikin.net | Dial-up UUCP client | boboto.cs-net.org | isc.cs-net.org | ista.cs-net.org D. Email flows for unikin.net and cs-net.org INTERNET | SMTP | v uucp.nsrc.org | UUCP | over | UUCP over TCP/IP TCP/IP| +-----------+ | | | v | v 194.7.233.44 194.7.233.195 ||| || dial-up ||| dial-up || UUCP ||| PPP || vvv vv cs-net.org unikin.net clients clients E. Other educational institutions have asked to also use the system, and we are in the process of adding them. ISC (Institut Superieur Commercial) IPN (Institut Pedagogique National) ISTA (Institut Superieur de Technique Appliquee) Facultes Catholiques V. Due to its geographical position in Africa and its human and natural resources, the DR Congo should play a significant role in the future of African communications, which will be ultimately dominated by the Internet. Working in a team and using the Internet as mean of communication between three continents has proven to be a fruitful way to bring the Congo Scientific Network from a vision to reality. We have taken the first steps to start an academic network in the DR Congo, with lots of volunteer help and efficient use of the tools of the Internet. Many things remain to be done : - CS-NET still operates based on a volunteer ethic by using people who are willing to help. A more effective way should be to get a minimal permanent staff. - ISP for universities: To be more effective, the scientific community of in the Congo should establish full TCP/IP connectivity to its universities and research centers. - We need to provide more support to Congolese researchers and students by organizing more research seminars and workshops with the participation of scientists from overseas who are actively involved in specific areas of research. Any suggestions on how to further develop our network, and improve our abilities to participate in collaborative scientific research with US scientists would be greatly appreciated. Antoine Bigomokero Bagula Congo Scientific Network ############################################################################### BIGOMOKERO ANTOINE BAGULA Home : Gardenia 2 Dept of Computer Science Skyvuedrive University of Stellenbosch Kluis Rivier Private Bag XI 7580 7602 Matieland Cape Town South Africa South Africa Email :bagula@cs.sun.ac.za bagula@unikin.net http://www.cs.sun.ac.za/projects/COE/#members Fax : +2721 8084416 (W) Tel : +2721 8084605 (W) Cell :+2782 9760453 ###############################################################################