Technical Aspects of Internet Governance: DNS, IXPs and Basic Infrastructure

It’s easy to realize that a single cable can be used not only for the Internet, but also for the TV and phone. That’s because the Internet uses the same structure as more traditional communication means. The difference, however, is that the Internet utilizes that structure in an innovative way, as to substitute these traditional means with more efficiency.
One of the aspects that allow for a greater efficiency is the manner of communication through which the Internet works. Inside the network, the devices communicate through an IP (internet protocol). The IP has two functions: to serve as an identification of a device on the network (through numbers separated in four series, such as 192.168.1.1) and to divide the information in different parts (packet).
IP packets can be described as part of an information, and each part is labeled with the IP addresses of its origin and the destination. The IP allows for networks with different technologies (VoIP; e-mail; Web; HTTP; pop; wi-fi; radio; etc) to interconnect on the same communication. It doesn’t matter whether you use cable, radios or fibers: all of them can be connected through the same protocol.
The Internet is a set of interconnected networks (inter for between, net for networks). That way, the IP is fundamental, because it allows for different functioning networks to connect with each other. The IPs, therefore, flow freely through these networks until its final destination.
Various persons (in a legal sense) integrate the Internet independently. The access providers (such as NET, Oi, GVT) allow for the final user to connect to the global network, granting a certain bandwidth in the circulation of information. These providers also have their own providers (transit providers), which, in turn, connect them to other networks and so on.
Google, Facebook and Twitter are in one of those networks and are called content or application providers. Therefore, in the Internet there are “wholesale” (transit) providers, which distribute their structures to the “retail” (access) providers. When these transit providers become too big, interconnecting various networks, they’re called backbones.
The various networks of the Internet (formed by access, transit and content providers), communicate with each other in order to know which is the fastest way for an information to travel through until its final destination. These networks are designed as autonomous systems (AS). In 2014, there were more than 50,000 ASes, with more than 500,000 routes. The relation between these ASes is commercial, so a provider can permit the information of an AS to flow through such provider in order to reach the rest of the network in a more efficient way, or in a faster manner. There are, also, the IXPs: an unique spot in which various ASes connect, which allows for an even greater speed in the exchange of information between them.
In order to access a content or application provider through a web browser, we don’t have to type in the IP of this providers’ server. This is due to the Domain Name System (DNS). The DNS works as a big chart which indicates a name in letters and its correspondence as an IP address in numbers. The DNS is actually distributed among various servers.
The root servers – of which the main one is in the US though there are a dozen copies distributed amongst several countries, Brazil not included – know where the first level domains are, which might be related to a specific country (.br, .pt, .tv, etc) or might be generic, which don’t relate to any country in particular (.com, .org, .info, etc). First level domains take information to another server, designed as authoritative. This DNS server knows the location of the server responsible for the hosting the web site we want to access. Our devices also contain a DNS server, called resolver.
To clear it up, I’ll exemplify the route of an information: when you type www.globoesporte.com.br → the information goes to the DNS system of your computer (the resolver) → which, in turn, sends the information to the recursive system made available by your access provider → this system, then, can do one out of two things: if the site was already accessed through it, it can remember the answer; if it doesn’t know where the authoritative server to that specific site is, the recursive system then sends the information to one of the nearest root servers → the root server  knows that the .br is in Brazil, but it doesn’t know where the authoritative server of the site is → the root server then sends information to the authoritative server of the .br → which doesn’t know in which server globoesporte.com.br is at, but does know where the site’s authoritative is → at last, the information is sent to the site’s authoritative (.com.br), which knows which IP is responsible for hosting Globo Esporte’s website. After completing this entire route the recursive server in your device allows the web browser to exchange information with that site’s server.