Domain Names and IP Addresses

Overview

The system involves two types of addressing: Uniform (or Universal) Resource Locators (URLs) and Internet Protocol (IP) addresses. The two types are linked by means of the Domain Name Servers (DNSs).

• URLs are mainly for users of the Internet, so they have letters and symbols that make them easier for people to understand, like:
  http://www.eblast.com/index.html
  Every file that's available on the World Wide Web must have a URL that uniquely identifies it - in other words, no two files in the world can have the same URL. A file's URL is a little like a person's name: everyone has one, and if you want to locate it, you need to know the person's name or the file's URL. Obviously, there's an important difference, too...people's names are not unique!
• IP Addresses are mainly for computers and technicians. Computers are most comfortable (efficient) with numbers, so IP addresses are composed entirely of numbers, like:
  198.111.176.122
  In fact, they work very much like phone numbers. Just as phone numbers are necessary if you want to talk to someone on the phone, IP addresses are necessary if you want to send information over the Internet. Fortunately, when we use the Internet, we don't have to figure out the IP address of every file we want, or every person we need to send email to. That's because of the DNS system...
• The DNS system's purpose is to connect the URLs with IP addresses. This is very much like the phone book: if you know someone's name and want to call them, you can look their number up in the phone book and dial the number. If your browser or email program has a URL or email address, it needs the corresponding IP address. The good news is that Internet computers take care of this pretty much on their own: they will ask a DNS server for the IP address and "dial" for you. As an Internet user, you should know in general how the system works, but you probably won't need to access the DNS system directly. Internet professionals, on the other hand, need to be somewhat more familiar with the DNS system.

 

Uniform Resource Locators		About that name: When it was originally proposed, the naming system used URL to mean Universal Resource Locator. Before long, people began saying that Universal is too ambitious (after all, the Earth is a very small part of a very big universe!), and suggested Uniform as being a more realistic name that preserves the same acronym. URLs are a subset of Uniform Resource Identifiers: URI. In fact, most URIs are URLs, but joining them are the Uniform Resource Names (URNs). A URN is a way of identifying a resource permenently, regardless of whether its physical location moves. Many Internet authorities prefer to use the more general term URI, rather than the more limited URL. (For more information, see W3C's Naming and Addressing information page, http://www.w3.org/Addressing/. Figure 1 illustrates the relationship between URIs, URLs, and URNs. URLs (and the Domain Name System in general) are organized hierarchically, very much like the file system on a computer. That is, there's a structure holding a large amount of "stuff", and that's subdivided into smaller and smaller bundles. The reason in both cases is the same: too much "stuff" to handle all in one bundle. On disks, we parcel up the "stuff" into folders or subdirectories.  On the Internet, it's parceled up into domains and subdomains. This allows information about servers and their addresses to be decentralized, so that not just one organization is responsible for assigning domain names and keeping the database up to date. If this hierarchical plan had not been followed, the explosive growth of the Internet would have been choked off or become unmanageable. Credits: The person responsible for designing this system is Dr. Paul V. Mockapetris, Program Officer for Networking at ARPA, later at Information Sciences Institute, University of Southern California. He is now (September 2004) Chairman of the Board and Chief Science Officer at Nominum (http://www.nominum.com/bio.php?id=2).
Anatomy of a URL		Let's take a look at some URLs and see how they're put together...   Protocol URL HTTP http://host.domain-name:port/path/filename.type#section FTP ftp://host.domain-name:port/path/filename.type File file:///device|path/filename.type Gopher  gopher://host.domain-name:port/gopherselectorstring WAIS  wais://host.domain-name:port/database?searchterm Telnet telnet://user@host.domain-name:port Mailto mailto:user@host.domain-name SMTP smtp:user@host.domain-name News news:news.group Thanks to Rasmuson Computing and Communications, University of Alaska / Fairbanks for the basis of this chart. The format (syntax) of a URL depends on what protocol is needed to handle the resource. So the first part of the URL is the protocol name.

Protocols		What is the Protocol, then? The Internet is the "information highway" for many different "vehicles". Each of these "vehicles" works in its own way, using its own internal rules. As long as these rules obey the general "rules of the road" (TCP/IP), these different "vehicles" can share the "road" very well. The protocol is the set of internal rules used for handling information in each of the various Internet "vehicles": HTTP: HyperText Transfer Protocol - "vehicle" for the World Wide Web FTP: File Transfer Protocol - for rapid transportation of files File: designates files on the users own (local) computer; the "rules" are dependent on the operating system of the computer they are on Gopher: the Gopher information system (precursor of the World Wide, now largely obsolete) WAIS: Wide Area Information System, a pre-Web system for searching out facts Telnet: for connecting to a computer and typing in commands directly SMTP (Simple Mail Transfer Protocol), Mailto: for sending electronic mail News: for accessing news servers
Top Level Domains		The National Telecommunications and Information Administration explains domain names this way: "The domain name space is constructed as a hierarchy. It is divided into top-level domains (TLDs), with each TLD then divided into second-level domains (SLDs), and so on. More than 200 national, or country-code, TLDs (ccTLDs) are administered by their corresponding governments or by private entities with the appropriate national government's acquiescence. A small set of gTLDs [generic top level domains] do not carry any national identifier, but denote the intended function of that portion of the domain space. For example, .com was established for commercial users, .org for not-for-profit organizations, and .net for network service providers." (National Telecommunications and Information Administration, Department of Commerce: DNS Statement of Policy 06-05-98, emphasis added) Most of the protocols share a common way of spelling out the domain name. Here are some examples:  umcu.org aol.com whitehouse.gov navy.mil umich.edu provide.net washtenaw.cc.mi.us uni-c.dk ukc.ac.uk The parts of domain names are separated by dots. The top-level domain (TDL) is on the right, usually a short (2-3 letter) abbreviation. Top-level domain names designate either what kind of organization owns the server (gTDLs), or what country it is in (ccTDLs).  Generic Domain Names (gTLDs) Code Meaning Example aero Aviation www.nw.aero biz Business www.bonbontree.biz com Commercial compuserve.com coop Cooperative organization nbca.coop edu Educational umich.edu gov Governmental whitehouse.gov info Information provider microbes.info int International organization paho.int mil Military ddn.mil museum Museum getty.art.museum name Personal names johnsmith.name net Network service provider provide.net org non-profit Organization ieee.org pro Professionals (doctors, dentists, lawyers, etc.) johnsmithdds.pro A lively debate is taking place about adding new domain names. For example, in 2004 ICANN tackled the issue of adding a new domain, .mobi, for mobile sites and mobile-oriented companies. What do you think about this? Is it helpful or confusing to add new, specialized domain names? After thinking and discussing it, read what Tim Berners-Lee, inventor of the World Wide Web, thinks about this proposal (http://www.w3.org/DesignIssues/TLD). Country Code Domain Names  (ccTLDs) Every country has a two-letter ISO abbreviation. Here are some commonly used ones:   Code Meaning Example Email Address au Australia psmith@deakin.edu.au ca Canada psmith@uottawa.ca de Germany katherine.krieg@student.uni-tuebingen.de dk Denmark karkov@sdf.uni-c.dk fr France geraldine@citilille.fr jp Japan hashimoto@keio.ac.jp nl Netherlands (Holland) vanthull@tue.nl nz New Zealand psmith@vuw.ac.nz se Sweden svensen@kth.se uk United Kingdom (England, Scotland, Wales...) jrm@ukc.ac.uk us United States krieg@wccnet.edu za Zud Afrika (Union of South Africa) psmith@und.ac.za For a complete list of country codes, see the Network Wizards Web site, http://www.nw.com/zone/iso-country-codes.  Subdomains exist within some of the geographical domains. For example, in the United States there is a subdomain for each state of the Union; within the states, certain types of organizations are given subdomains as well. Examples: washtenaw.cc.mi.us ci.ann-arbor.mi.us co.washtenaw.mi.us where mi (Michigan) has a cc (community college), a ci (city government), and a co (county government) subdomain.. Some, but not all, other countries assign subdomains for academic, commercial, non-profit, and so forth. Examples in the chart above include domains au, jp, nz, uk, and za.
Second-Level Domain Names (SLDs)		Let's take a look at the example names again: umcu.org  aol.com  whitehouse.gov  navy.mil  umich.edu  provide.net  washtenaw.cc.mi.us  uni-c.dk  ukc.ac.uk In each of these names, the left-most name represents the domain name. It is either an abbreviation of the organization's name, or a spelling-out of the name. The idea is to make the domain name easy for people to associate with its organization (or commercial product).

Server (Host) Names		In addition to a domain name, a host name may be present. The host is the server (computer) that will actually be doing the work - sending or receiving a file. Here are some examples, with the server name in bold: ccmail.lfa.com gala.wccnet.org www.wccnet.org  fc.skolekom.dk ann-arbor.va.gov student.uni-tuebingen.de sdf.uni-c.dk orchard.washtenaw.cc.mi.us med.va.gov The host name is optional. It may or may not be used depending on how the domain's servers are organized.  If there is only one server in an organization, no host names are needed, but they may be used anyway.  Many organizations with multiple servers, such as Washtenaw Community College, enter each of the servers with a different host name in the Domain Name System (see below) so computers outside the organization can address each server individually. Other organizations with multiple servers, such as America Online and the University of Michigan, have their own automatic system for deciding which of their servers should respond to a request or receive an incoming message. That makes their email addresses simpler to remember and to type.
Email addresses		The electronic mail protocols (SMTP, mailto, POP, IMAP, etc.) expect the individual user's name to the left of the domain name, separated by the @ sign. Here is an example email address: psmith@ruby.xyz.com psmith is the user name. Often a combination of first and last name or initials, but may be a number assigned by the system administrators. ruby is the host name.  xyz is the domain name.  com is the top-level domain name.
World Wide Web URLs		The HTTP protocol says that information more specific than the domain and host name should go to their right. Let's take a few apart... http://www.ipivot.com/index.html  http://zorba.uafadm.alaska.edu/DCC/training/  http://www.ipivot.com/about/about-pmockapetris.html http://www.wccnet.org/dept/cis/mod/ng04c.htm#TopLevel http://websearch.about.com/internet/websearch/library/weekly/aa101999.htm?iam=dp&rf=dp  http identifies to the protocol used by that address Between the slashes, as in /www.wccnet.org/ is the domain and server name.  The path name follows the server name: about/about-pmockapetris.html. URL path names are standardized using the Unix-style forward slashes to separate directories or folder names. The file names also follow Unix naming conventions, usually with a dot-separated extension giving the file's type. If there is no file name, the HTTP protocol specifies that the server search for and return either: index.html default.html index.htm default.htm a listing of the files in the directory, if permission has been granted to display such a listing. A double-cross # (pound-sign) can be used to specify an anchor: ng04c.htm#TopLevel. This allows pointing not only to a file, but to any number of locations within a file. A question mark ? indicates the beginning of a CGI (Common Gateway Interface) string such as aa101999.htm?iam=dp&rf=dp. CGI is a protocol that allows the user (client) to send a string of characters for the server to process. Often, these are returned when the user fills out a form or types in a word to be searched for. In order to work, the server must have a program which will process the string and respond to the client by sending HTML code for display to the user.

Internet Protocol Addresses		IP addresses are the numbers actually used to identify devices attached to the Internet. Their role is very much like that of telephone numbers in the phone system: They are used to electronically set switches in routing equipment, allowing signals to be sent to the desired destinations. The most widely used standard for IP addresses is version 4, referred to as IPv4. Each address is a 32-bit binary number, usually written as a series of 4 base 10 numbers separated by dots:  12.34.56.78 Each of the four sets of numbers represents 8 bits, which can range between 0 and 255. The lowest IP address is 0.0.0.0 and the largest is 255.255.255.255. Using this system, it is possible to represent 4,294,967,295 distinct addresses.  Although having more than 4 billion possible addresses seemed quite enough twenty years ago, we are rapidly running out of IP addresses. In order to take care of the problem, the IETF has designed a new standard, IPv6. (What about IPv5? It was an experiment in using the Internet to stream packets for high-speed applications like video; it was intended to work with IPv4, but was eclipsed by proprietary streaming technologies like RealMedia.) As of February, 2003, IPv6 is working smoothly with IPv4. There are applications - both client and server oriented - currently available. The expectation is that IPv4 will gradually fade away as more Internet hardware and software are upgraded. For more information, visit the IPv6 Forum Page: http://www.ipv6forum.com/
IP Address Classes		Under IPv4, organizations that distribute IP addresses give out "chunks" of addresses to user organizations. These "chunks" come in three sizes, or classes depending on the needs of the organization and the number of addresses available: The largest number of addresses Intermediate Smallest

 

Domain Name Servers		Domain Name Servers (DNS) are computers with a database associating each domain name with an IP address. Whenever an email is sent out or a browser requests a Web page, the DNS is asked what that URL's IP address is. This is like calling telephone information service to ask for a person's phone number: you can't get through with just the name - you need the number.
		Because domain name lookup is so frequent and so vital to moving information on the Internet, there are lots of domain name servers. But because IP addresses are added and changed so frequently, there is a need for continual update. This gave rise to a system of registering DNS computers and sending frequent updates around the system. The process is complicated by the fact that there is no one central authority on the Internet that assigns all the domain names to IP addresses. Instead, updated information is circulated by multiple domain name authorities, and all DNS computers modify their listings according to these updates.
References:		For more information about the Domain Name System: Domain Name System Structure and Delegation (Jon Postel): http://www.isi.edu/in-notes/rfc1591.txt The US Domain: RFC 1480 (Jon Postel and Ann Westine Cooper): http://www.isi.edu/in-notes/rfc1480.txt

Review		Click here to see the review questions and answers for this module.
Audience		This is for people who are generally familiar with computers and the Internet, and want to learn more about how Internet addressing works. See "What is the Internet?" (module NG03c) for more background.
Objectives		When you successfully complete this lesson, you will be able to... Explain the domain name system and commonly seen top-level domain names; current efforts to expand top-level domain names Explain IP addresses; current plans to expand IP address space Define URL Identify URL's purpose Explain components of URLs Identify types of URLs Explain the Domain Name Server system
About this Document		Module NG04c: Domain Names and IP Addresses This document is part of a modular instruction series at Washtenaw Community College. For more information, see the overview or the list of modules in this series, NG "Networking: General Overviews and Getting Started." This document has been used in the following classes: INP 159, INP 160, INP 220. Author: Laurence J. Krieg Institution: Internet Professional Department, Washtenaw Community College History: Original: 24 Nov 1999; this version Sunday, 16-Jan-2005 15:55:52 EST Copyright: Copyright © 1999-2005, Laurence J. Krieg. Instructors: You may point to this file in your Web-based materials.  Students: you may make a copy for your personal use.  All other uses: contact the author, Laurence J. Krieg for permission.