The Truth About Servers

When people refer to a piece of a hardware as “a server,” they typically mean that it is running one or more pieces of server software, may or may not be dedicated to that role, and is possibly made up of higher-grade components that tolerate long periods of availability.

When people refer to a piece of a hardware as “a server,” they typically mean that it is running one or more pieces of server software. You must go under the hood to really understand what the server is doing.

Of course, in either case, appropriate software is the core of the system. When people refer to a piece of a hardware as “a server,” they typically mean that it is running one or more pieces of server software, may or may not be dedicated to that role, and is possibly made up of higher-grade components that tolerate long periods of availability.

Requests are sent to a server by a “client.” A good example of a client is the garden variety Web browser typically used as client to a Web server. Networks are the backbone of client/server relationships, and the rise of the Internet and local area networking (Intranets) has seen the evolution of a wide variety of servers.

Many servers in use today are rooted in historical models that have been a part of computing for years; others are evolutions of these models, often tagged with new, marketing-friendly names. The upside of this market flood is an atmosphere of competition and improvement; the downside can be confusion about category names that are sometimes invented to help differentiate a product from its closest peers and competitors.

In this tutorial, we’ll survey the wide variety of servers available, from familiar Web servers to so-called “application servers”, proxy servers, e-mail servers, as well as DHCP servers, firewalls, and fax servers. We’ve broken the server types to be discussed into four categories: Web servers, “children” of Web servers, intranet-level servers, and on-demand Internet servers.