There are a significant number of different servers that were specifically created to run Java applications on them. This section will provide a brief orientation on them so that you can more easily choose which server to use for your application.

All Java servers can be roughly classified into two groups:

The main component of the first group of servers is the servlet container. They can have other specifications and services implemented too, but all their functionality is centered upon hosting Java web applications.

The most famous and widely used representative of the first group is Apache Tomcat. The other free server, especially popular among Tapestry developers is Jetty. Gaucho Resin, yet another representative of the first group, is not free but has a very good reputation.

If all you need is a web application, even if it uses popular frameworks like Spring or Hibernate and a database to store its information, any Java web server should be fine for you. If, however, you decide one day to use Enterprise JavaBeans or Java Message Service, you will most probably need a full Java EE application server.

The second group implements a wide range of Java EE specifications which means that the purpose of these servers is to host multi-tier, distributed enterprise-level applications that communicate across the network, coordinate the work of their parts and can withstand a significant workload.

A few representatives of the second group are JBoss, Apache Geronimo, IBM WebSphere or BEA WebLogic. Some of them are free, others are very expensive, but in any case I would suggest that these servers should not be used unless you really need the functionality provided by them. As a side note, several full Java EE servers, like JBoss and Geronimo, use Tomcat for their servlet container.

This completes a brief overview of Java EE-related terminology. I hope you found it useful, especially if you are new to Java web