Let your fingers find the web sites Domain Name System is able to pick the specific page you want out of millions of sites located a round the world by DANA H. SHULTZ

Have you ever wondered how your Web browser is able to pick the specific page you want out of millions of sites located throughout the world? The key is the Domain Name System.

Human beings refer to Web sites by their Uniform Resource Locators (URLs). For example, Microsoft's Web site is named www.microsoft.com.

The Internet, in contrast, refers to sites by their Internet Protocol (IP) addresses. Every computer connected to the Internet has a unique IP address.

IP addresses consist of four numbers separated by periods. For example, the IP address for the www.microsoft.com server is 198.105.232.4. The first number identifies the geographic region, the second specifies the organization, the third indicates a group of computers and the fourth uniquely denotes the specific server.

Providing translation

When you enter the URL into your Web browser (probably Netscape Navigator or Microsoft Internet Explorer), the URL is transmitted to your Internet service provider (ISP). The ISP's Domain Name System (DNS) server is responsible for translating the URL into the corresponding IP address.

The DNS server has a data base that stores the IP address for every URL it has processed in the past. If the URL you entered is in this data base, the DNS server finds the right IP address and connects your browser to the Web site.

If the URL is not in the DNS server's data base, the DNS server asks an "authoritative" DNS server for help. An authoritative DNS server can state, with certainty, whether or not your URL exists in the "domain" that the DNS server covers.

Let's digress for a moment to explore how domains are structured. (Caveat: This digression is not for the technically timid.)

Domain hierarchy

To facilitate navigation, the Internet is divided (logically, not physically) into a hierarchical set of "domains." The "root" domain sits at the top of the hierarchy. (Only in the arcane world of the Internet could the root be at the top!) Information about this domain is stored on a limited number of authoritative "root servers" located at various places on the Internet.

Immediately below the root domain are the "top-level" domains. Within the U.S., top-level domains are associated with organization codes ("COM" for commercial organizations, "EDU" for educational institutions, etc.). Outside the U.S., top-level domains are associated with countries (such as "CA" for Canada).

Below the top-level domains are second-level domains, such as microsoft.com. Once a second-level domain is created (the Network Information Center must OK the domain name), the owner can create subdomains, such as www.microsoft.com, at will. For a subdomain to be accessible, information about it must be stored on a server in the immediately-higher domain.

Resolving the problem

Let's assume, now, that you are looking for www.fictitious.com. When your ISP's DNS server cannot find the URL, the DNS server contacts a root server.

The root server knows whether or not fictitious.com exists. If it does, the root server replies that information about www.fictitious.com can be found at fictitious.com. The DNS server goes to fictitious.com and finds the IP address necessary to connect you to www.fictitious.com. 

On the other hand, if fictitious.com does not exist, the root server so notifies the DNS server. You then receive a message saying that the URL you specified is unknown. Whatever the result, the real miracle is that this Rube Goldberg method of communication works at all!

Dana H. Shultz is an Oakland lawyer, certified management consultant and speaker specializing in office technology and online marketing. He may be reached by e-mail at dhshultz@ds-a.com and on the World Wide Web at http://seamless.com/ds/.