Encyclopedia of Database Systems

Living Edition
| Editors: Ling Liu, M. Tamer Özsu

Java Enterprise Edition

  • Ricardo Jiménez-PerisEmail author
  • Marta Patiño-Martínez
Living reference work entry
DOI: https://doi.org/10.1007/978-1-4899-7993-3_1542-2


Business Logic Server Tier Java Message Service Aspect Oriented Programming Enterprise Java Bean 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.



Java EE (JEE) is a Java Community Process (JCP) specification for Java application servers. JEE consists of a set of related specifications and APIs that shapes an ecosystem for building Java distributed applications over a multi-tier architecture, including web, business logic, and storage tiers. JEE provides components specific for each tier. JEE provides servlets and Java Server Pages (JSPs) for the web tier, Enterprise Java Beans (EJBs) for the business logic tier, and a data driver (JDBC) and an object-relational mapping (via entity beans) for the storage tier. Components can be distributed at different sites and interact with each other. The interaction among distributed JEE components is achieved via RMI (Remote Method Invocation), the JEE specification for remote procedure calls. In its latest editions, JEE has been smoothly integrated with XML processing and web services. It is possible to program web services using EJBs and a large set of APIs enable a standard way to effectively manipulate XML. JEE covers all important non-functional aspects such as transactions, security, etc. Transactions are core to JEE and are considered across all tiers in a consistent fashion through the Java Transaction Service and API (JTS/JTA). A number of specifications address various security issues such as authentication (JAAS, Java Authentication and Authorization Service), encryption, and so on.

Key Points

The JEE project originated back in 1998 when the JPE (the original name) initiative was announced by Sun Microsystems. J2EE 1.2 was released in Dec. 1999. The following versions have been released almost every 2 years, J2EE 1.3 in Sept. 2001 (1.3 was the first version to be developed as a JCP specification), J2EE 1.4 in Nov. 2003, JEE 1.5 (with version 1.5 J2EE was renamed as JEE) in May 2006, and JEE 1.6 is expected to be released sometime in 2008.

JEE is a multi-tier middleware framework to program all kinds of applications based on the Java language. Multi-tier architectures have become widely used since they allow an adequate separation of concerns. Each tier has a specific container with a component model tailored to the specific mission of the tier. Some of the most common tiers are: web tier, business logic tier, and persistent storage tier.

Figure 1 depicts a typical JEE multi-tier architecture. Clients interact with the system via the web server tier. The web server tier then delegates the business logic to the application server tier. The JEE application server tier can consist of multiple servers, that is, be distributed. Then, the persistent data are kept in the database tier. The data can be kept in multiple distributed databases. The database instances can be shared across application servers. Transactions are used to enforce data consistency across the entire system despite concurrent accesses to the data. Transactions also protect data against server failures providing failure atomicity.
Fig. 1

JEE multi-tier architecture

For the web tier, JEE provides a component model for generating dynamic web content. This component model is based on Java Server Pages (JSPs) and servlets. JSPs provide a high level abstraction for dynamic content, whilst servlets provide a lower level view. In fact, JSPs are translated into servlets. Servlet containers are specialized web servers that support servlet execution. Typically servlets deal with the presentation aspects of an application and they are specialized to render dynamic web contents to users. Servlets delegate the business logic to the application server tier, that is, to the JEE application server.

The JEE application server provides a component model for writing general applications. The components in the JEE component model are termed Enterprise Java Beans (EJBs). Two different kinds of EJBs are distinguished: session beans and entity beans. Session beans enable to encapsulate the interaction with users. There are two flavors: stateless and stateful session beans. Stateless session beans provide support for non-conversational interactions, whilst stateful session beans support conversational interactions. Stateless session beans are instantiated per client invocation and their state is discarded after processing the client request that triggered their instantiation. On the other hand, stateful session beans are created upon the first client invocation and then, they are kept during all the client conversation. Their state enables to tracking of the status of the interaction across invocations (e.g., a shopping cart). From JEE 1.5, it is also possible to persist POJOs (Plain Old Java Object, that is, a Java Object) using JPA (Java Persistent API) without the need of a JEE application server.

JEE provides two models of interaction with the database: explicit and implicit. In the explicit model of interaction the EJBs access the database directly via a JDBC driver. The JDBC driver abstracts the database specifics of the underlying implementation providing a uniform vendor-independent interface. However, the object oriented paradigm has a different style than the relational model (the well-known object-relational impedance mismatch) that forces programmers to write a lot of repetitive code prone to errors to create queries and to transform query results into objects. The object-relational mapping (ORM) approach addresses this impedance mismatch by offering the programmer a view of an object oriented database, eliminating the need to write code.

The database tier access can be explicit via JDBC commands or implicit by using the object-relation mapping (ORM) of JEE. Explicit access implies issuing JDBC commands with SQL statements and dealing directly with the results sets returned by JDBC. Implicit access via ORM enables access to the persistent state using the object-oriented paradigm. Tuples are modeled as special kind of JEE components, entity beans, that are accessed as regular objects. Entity beans provide an object-oriented cache of the database. The JEE system takes care of updating the database and implementing the necessary concurrency control to provide transactional serializability.

JEE also provides support for the publish-subscribe paradigm. For that purpose it provides a specific component model and the associated container, Java Message Service (JMS). JMS provides message-driven beans that are components that are activated upon the reception of a particular kind of message. Message-driven beans can be invoked by client applications and EJBs and they can also invoke other EJBs. JMS provides an asynchronous interaction model that complements the synchronous one provided by EJBs and RMI.

In the last few years, a number of initiatives have tried to simplify JEE development. One of the initiatives with highest impact has been the Spring framework [2]. Spring provides a lightweight approach to JEE simplifying the development of JEE applications. Spring is based heavily on programming interfaces (as opposed to programming objects) and the use of aspect oriented programming. Spring has inspired some of the major changes in JEE 1.5 and 1.6.


Recommended Reading

  1. 1.
    JSR 316: JavaTM Platform, Enterprise Edition 6 (Java EE 6) Specification. http://jcp.org/en/jsr/detail?id=316
  2. 2.
    Spring Framework. http://springframework.org/

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  • Ricardo Jiménez-Peris
    • 1
    Email author
  • Marta Patiño-Martínez
    • 1
  1. 1.Universidad Polytecnica de MadridMadridSpain

Section editors and affiliations

  • Cristiana Amza
    • 1
  1. 1.Dept. of Elec. and Comp. Eng.Univ. of TorontoTorontoCanada