Tuesday, March 6, 2012

Of Gateways and Gatekeepers: The History of Internet Exchange Points in Kenya and Rwanda

An excerpt from "At the Crossroads: ICT Policy Making in East AFrica" ©International Development Research Centre 2005, First Published 2005, ISBN 9966-25-439-0

Note: This piece chronicles the role I played in setting up KIXP between 2000 and 2002


The Internet in Africa has been growing steadily over the past several years and is beginning to play a significant role in Africa's development, creating employment, providing opportunities for innovation and entrepreneurship, as well as acting as an enabler in the digital delivery of government services, education, radio, and healthcare among others. The new possibilities provided by Internet technologies present African countries with an opportunity to leapfrog phases of development and make use of the most recent innovations to establish a strong information society and increase the distribution of wealth among the populace, thereby addressing the poverty that has plagued the continent to date.

Unfortunately, the overall impact of the Internet as an enabler in Africa has been severely curtailed by a number of elements. Topping the list is the lack of efficient paths to carry growing local and regional traffic between Internet Service Providers (ISPs) in Africa. This problem occurs both on a national as well as on a regional or inter-country scale. For example, when an African Internet user sends a message to a friend in the same city or a nearby country, that data travels first to Europe or the United States before getting back to the African city or nearby country. It has been estimated that this use of international bandwidth for national or regional data costs Africa in the order of US$400 million each year. This situation has its parallel in telephony where it may be easier to route a call from say Nairobi or Dar es Salaam via Europe or the United States to a neighbouring country than to do so directly. A connected Africa is vital. When African countries communicate with each other in this manner, the loss of money due to inefficient traffic routing is astounding. The net result is that poor African countries get poorer.
The vision of a connected Africa begins with the building blocks of the Internet. At the national level, Internet traffic between ISPs has been optimised in a number of countries with the introduction of Internet Exchange Points (IXPs), which allow ISPs to interconnect and offload correspondent traffic. Only 11 of the 53 countries in Africa have internet exchange points. The result is an inefficient exchange of African inter-country traffic through hubs located overseas mainly in the US and Europe. This means that Africa is paying overseas carriers to exchange "local" (continental) traffic on its behalf. This is costly and inefficient. It is, therefore, well within the interest of all countries in Africa to find ways of optimising Internet traffic, through building better and more robust networks to support intra-continental traffic flows. This will create opportunities for private sector investment.


For most African countries, access to the Internet began with the national telecommunications corporation offering digital links, mostly satellite based, to the local market which consisted mostly of ISPs and other Internet access providers such as cyber cafes or telecentres. In essence, the national telecommunications corporation became country 'gateways' to the Internet and, by extension, the rest of the world. Herein lies the origin of the 'gateway-centred' Internet development approach in Africa, and the gate-keeper phenomenon. This genesis for the Internet meant that as the local demands for Internet access grew, so did the size of the "gate." 'Gateway-centred' Internet development evolved. Internet access is measured in terms of bandwidth, which is "the capacity for data transfer of an electronic communications system." This meant that the local "gateway" increased as the amount of bandwidth available to the local market. This was arguably the most logical form of growth, until a myriad of problems revealed the inherent weaknesses. With the 'gateway-centred' Internet development, almost overnight, the very facilities, which had been touted as gems in the communications industry became the biggest nightmare. This can be clearly seen in the example of Kenya's 3-day Jambonet network failure in 1999. Despite the fragile situation presented by a country depending on one single gateway for all of its Internet communications, the national telecommunications companies fought hard to keep the market from being "opened up" to competition. In time, the very people supposed to facilitate access and connectivity for the country became the bottleneck, the gatekeepers, locked up the country's communications sector tightly.
The budding ISP community in many African countries became agents of change and quite often the most vocal on issues relating to sector liberalisation. Among some of the initiatives triggered by ISPs, the most significant was the introduction of IXPs into the world of African Internet. Cisco Systems, the largest networking company in the world has said "IXPs are the keystone of the entire Internet economy". IXPs are the interconnection points of the Internet. In other words, IXPs are the places where ISPs interconnect with each other. An IXP can be described as a clearinghouse for Internet traffic, and the mantra among IXP practitioners is "Keep local traffic local."
The majority of ISPs use their international circuits to carry traffic to other ISPs within the same country. In some cases, an email sent by a user to a correspondent across the street will be carried out over one ISP's international circuit, to the US or wherever the upstream provider's network terminates usually in Europe or the US, and then back down another ISP's international circuit to the intended recipient. This unnecessary routing of traffic consumes a lot of bandwidth on these international circuits and causes unnecessary congestion. Various reports suggest that between 30-40% of all Internet traffic is local.
The commercial nature of the Internet, and the rapidly growing traffic it has generated, has provided tremendous incentives for ISPs to increase the number of IXPs. The locations, names, and years of establishment of existing IXPs in Africa are as follows:
• South Africa – JINX, established 1997
• Kenya – The Kenya Internet Exchange Point (KIXP), established 2001
• Uganda – UIXP, established 2003
• Tanzania – TIX, established 2003
• Mozambique – MOZIX, established 2003
• Egypt – EGIX, established 2003
• Nigeria – IBIX, established 2003
• Democratic Republic of Congo – KINIX, established 2003
• Rwanda – RINEX, established 2004
• Swaziland – SZIX, established 2004
Despite the significant increase in momentum with the creation of IXPs over the past two years, it is worth noting that there is still a glaring need for IXPs on the continent. Despite this need, setting up IXPs has not been easy.


Localising Local Internet Traffic

The concept of a facility that would ensure local routing of inter-ISP traffic within Kenya was first mooted within the ISP Task Force of the East African Internet Association in 1997. At the time, ISPs in Kenya obtained their international Internet connectivity through international leased lines from the Kenya Posts & Telecommunications Corporation connected to European and American network service providers. At the time, KPTC was the only telecommunications organisation allowed by law to carry international traffic from the point of origin to termination.
In 1998, KPTC launched Jambonet, the country's Internet gateway and backbone. Jambonet provided local ISPs with in-country access to the Internet via high capacity links between KPTC and international network operators. The cost of international Internet connectivity was reduced by over 50% with the introduction of Jambonet services leading to the termination by most ISPs of their international leased line connections choosing instead to connect to Jambonet. However, despite links to the local gateway, most inter-ISP traffic was still leaving the country and returning after traversing one or two other countries. This instigated the drive for a local facility that would keep Kenyan traffic within Kenya because of the costs.
After attending a Networking Workshop for Developing Countries hosted by the Internet Society (ISOC) in San Jose, California, US, in June of 1999, one of Kenya's Internet engineers obtained knowledge about how to design, set up, and maintain an IXP. Upon returning to Kenya, he proceeded to share this information with other ISPs. This new information instigated and helped the young ISPs to crystallise their vision for a local IXP.
During the various discussions relating to the establishment of the IXP, a number of different models were evaluated for the Kenya IXP. It was agreed, on a technical level, that the KIXP would be based on the same model as the Hong Kong Internet Exchange. This model is called the Layer Two Route Reflector IXP Model.
Architecturally, the KIXP core is designed to transmit data at the fastest possible speed and consists of two high speed ethernet switches. Each IXP member has the option of connecting their routing equipment to both switches. That way if one switch should fail, the other would take over automatically. The core is supplemented by two 'route reflectors' – specially configured routers that will bounce routing logic to all members at the KIXP until there is convergence, and all the routers have the same view of the network. This design aspect allows for quick and easy policy implementation at the exchange point. The design enforces a multilateral peering policy for all members, thus allowing everyone to peer with everyone else. This design reduces the amount of configuration that needs to happen at the ISP router. The ISP then only has to have a peering session with the route reflector to see and exchange traffic with everybody else at the exchange point. Based on the existing design, the exchange point is capable of supporting up to 48 ISPs. This capacity can be extended further to support up to 200 individual ISPs.


Setting up the IXP involved debates on the location of the IXP, operational resource considerations, and an evaluation of the local traffic. One of the biggest debates was with regard to where the IXP would be hosted. A number of options were evaluated which included the following:
• The first option considered was Telkom Kenya, ostensibly the most suitable option since it was the incumbent public national telecoms operator. Some of the reasons cited in favour of Telkom Kenya included the fact that as national operator, all ISPs already had existing data links to its data network. Additionally, due to its central location, it would be much easier for all ISPs to gain physical access to the IXP, regardless of their location. This option, proved to be unworkable because Telkom Kenya declined the ISPs' request to host KIXP.
• The University of Nairobi was considered as an alternative host for KIXP mainly due to its dynamic computer studies faculty and its central location. The biggest concern about using the university was the frequency of student riots. Since the KIXP was expected to serve a mission critical purpose, this concern eliminated the university as a viable and serious option.
• A couple of ISPs that had their offices conveniently based in the CBD offered to host the IXP. The challenges here were both which ISP to choose out of the two as well as the fact that most of the other ISPs expressed a high level of dissatisfaction and would not trust them as competitors to handle the IXP without seeking to give themselves undue and possibly unethical advantage.
• After an evaluation all of the various options without finding one that satisfied all the ISPs or with which they were all equally comfortable, the idea of leasing space in a conveniently located building was posed. This idea was well received and allayed most of the fears and concerns expressed by the ISPs. It was finally unanimously decided that a neutral, leased facility would be the best. A lease was taken up for 1500 square feet on the top floor of a strategically located building in the Nairobi city centre. So began the KIXP.
A main operational resource consideration was cost. As with any other type of data networking or communications infrastructure, costs fell into two broad categories: setup and operating costs. Setup costs included the cost of purchasing equipment for the core of the IXP as well as furnishing the room where the IXP was to be located with backup power, air-conditioning, equipment cabinets, and the relevant security fixtures. The initial equipment was funded both by a donation from Cisco Systems Incorporated as well as a small grant from the United Kingdom's Department for International Development. Other setup expenses were covered by funds from the Telecommunications Service Providers of Kenya (TESPOK). Since the space where KIXP was located was not free, it was necessary to find a way of covering the operating costs, such as rent, electricity and insurance costs. A monthly subscription fee for all members connecting to KIXP was introduced to cater for operating costs.
The exchange point went live on the 21 October 2001 with four connected ISPs exchanging traffic initially. An assessment of the local traffic since the initial stages, apart from the period during which the KIXP was shut down due to hostile regulation, shows that there has been a significant impact and improvement on the performance of the local infrastructure.


The challenges that faced the IXP included an all out attack on the KIXP with the threat of a shutdown; a legal battle to get KIXP back on track; coming up with a compromise solution; and an operational policy review. The moment the announcement that KIXP was live went out and following press releases to the local dailies to the same effect, an almost immediate attack issued from Telkom Kenya, the national telecoms operator, and incumbent. According to Telkom Kenya, KIXP was operating illegally and contravening exclusivities granted to Telkom Kenya which were valid until 2004. Telkom Kenya, exercising political muscle, put a high level call through to CCK, the regulator, to shut down the KIXP. The CCK wrote to KIXP demanding immediate shutdown on the basis that the facility was operating without a licence, an order that was effected.
The CCK order and the ensuing legal battle to get KIXP back on track were received with much alarm by the ISP community. The ISP community was determined to fight the battle. Close scrutiny of the Communications Act '98 and the Regulations of 2001 revealed the startling truth that KIXP did not need a licence to operate as it was a facility operated cooperatively by licensed providers. A case was then presented to the Communications Appeals Tribunal with a strong technical argument showing that KIXP was merely a standard, off-the-shelf ethernet hub. If the KIXP were to be shut down, then the CCK should shut down every computer network in the country since the technical architecture and components were equivalent. The Telkom charge was thus shown to be invalid. Thus was the road to a compromise solution paved. The CCK's response to this appeal, duly served by the Communications Appeals Tribunal, was an informal phone call to the chairman of TESPOK, the ISP association, with a request to explore an out-of-court settlement. It was clear that Telkom Kenya had misrepresented the situation and, because the matter was made public and had received a significant amount of attention and coverage in the local and international media, it was necessary to find a face saving solution and make everybody happy.
The approach eventually adopted was the establishment of a company called KIXP Limited, which then applied for an IXP licence, which CCK duly granted. This made Kenya the first country in the world to have an IXP licence.
Regarding operational policy relevance, in the time that KIXP has been operational, it has become a key and integral part of Kenya's local Internet infrastructure and has made it possible to have robust, real-time, online applications which are relevant to the local community to be developed. A good example is the online employment web portal, www.myJobsEye.com which has thousands of jobs and jobseekers in its database and has been able to significantly shorten the employment/job seeking time.
In the course of 2004, it emerged that the policies that governed membership and use of KIXP were restrictive since they allowed only licensed ISPs to be members and to connect to the IXP. This realisation prompted a policy review which lifted all restrictions on membership and lowered joining fees by 600%. The new policy has allowed non-ISPs such as the academic network Kenya Education Network (KENET), as well as the national domain name registry, Kenya Network Information Centre (KENIC), to become members and localise their traffic.


PETER said...

I hope u r thinking of penning a book on this stuff one of this days for enthusiasts in the field... the history of connectivity in kenya.

Hosted PBX Providers said...

Thanks for sharing your info. I truly appreciate your efforts and I am waiting for your next post thank you
once again.