Introduction to Java Enterprise Edition

Introduction to Java Enterprise Edition

• 1.
  Introduction to JavaEnterprise Edition 1 By: Abdalla Mahmoud Contents Introduction to Java Enterprise Edition............................................... 1 Contents ........................................................................................... 1 Preface ............................................................................................. 3 1. Distributed Systems ........................................................................ 3 1.1. Application Architecture ............................................................. 4 1.2. Application Design ..................................................................... 4 2. Middleware .................................................................................... 4 3. Java Enterprise Edition Platform........................................................ 5 3.1. Application Architecture ............................................................. 6 3.2. Application Design ..................................................................... 7 4. The Java EE Application Server ......................................................... 9 4.1. Runtime ................................................................................... 9 4.2. Services.................................................................................. 10 4.3. Final Image - Java EE Architecture .............................................10 1. http://www.abdallamahmoud.com/ 1
• 2.
  2
• 3.
  Preface In Java SE,we used to develop desktop applications. Now imagine the situation where many desktops should be involved in a certain system; it's found in customer support centers, modern retailer shops, eGov systems, eLearning systems, and so on. Those types of systems consist of many computers connected together via a network, and all operating in a central database, cooperating to form a single and concrete system. These systems are called distributed systems. 1. Distributed Systems A distributed system is a collection of computers connected together via a network, and appear to the users as a single system. Different characteristics are common to all distributed systems, including resource sharing, openness, concurrency, scalability, fault tolerance, and transparency. The web is a common example for a distributed system. Resource Sharing: in a distributed system, resources are used by many users and computers in the network. So resources like files, printers, databases, and so on should be shared. Resource sharing includes naming each resource in the system, and provide it with a standardized interface to be accessed with. Openness: Openness describes the ability of interacting with others. We use openness term for describing one's ability of making communication with the others. It's the same in distributed systems, openness describes system's ability for interacting with other systems. Concurrency: many users will access the system simultaneously, use resources like files and databases in the same time. Concurrency is a property of a distributed system describing its ability for managing many users access to resources instantaneously. Scalability: the ability of a distributed systems to adding new resources and users. Fault Tolerance: computers fail! When a computer in a distributed system is down, the distributed system should continue working and functioning properly. Transparency: resources should be decoupled from the users. Users should be able to access resources transparently without caring where the resource is actually located in. Many users should access the same resource without taking in concern that many other users may access the resource in the same time. In addition, resources may be moved from a location to other while in use, without stopping the system or notifying the users, where the component should continue working and functioning properly. This's called transparency. We're going to study distributed systems from two prescriptives: • Application architecture. • Application design. 3
• 4.
  1.1. Application Architecture Adistributed system is a collection of computers connected together via a network. We have studied client/server architecture (where the application is divided into two tiers: server tier that provides a service, and a client tier that consumes the service) in two contexts: socket programming in the ShoutBox application, and database programming when we used JDBC API together with the JDBC driver to connect to a relational database server (PostgreSQL). Most distributed systems are implemented in what's called Three-Tier architecture. Three- Tier architecture is an extension to the client/server architecture where business data, application processing, and presentation are separated from each other. A common example is separating Database (business data), business logic (application processing), and user interface (presentation) from each other. In other words, the database is an independent subsystem, the business logic is an independent subsystem, and the user interface is an independent subsystem, as well. Some issues should be concerned when designing an application with such architectures: Protocols as computers connect together, protocols should be defined. Most protocols are too complex to implement and abstractions are needed. Security as computers connect together, they are exposed to each others. Only authenticated users should access the systems, and only authorized users should be able to access resources. In addition, protocols may provide some type of security by encrypting conversation using encryption techniques like public/private key. 1.2. Application Design Most distributed systems are designed in what's called component-based model. The system is composed of well-defined software components that are assembled together to form the final system. The definition may seem similar to the object-oriented concept, but they differ. A software component is a logical term describing a self-contained software unit that's accessible using standard interfaces. A software component is different to software objects in that software components are accessible within the scope of a distributed systems (i.e. components can be accessed remotely), where software objects are accessible only within the scope of single application (i.e. objects cannot be accessed remotely). 2. Middleware As we just seen, distributed systems require heavy system programming. It's too complex to build a distributed system; it requires too complex implementaiton of the system architecture, communication protocols, t architecture, support to components, etc. In addition, the purpose of the system target mainly the business itself (not the distributed system), developers should focus on implementing the business logic itself, not the complex distributed system requirements. Hence, the term middleware has been introduced. A middleware is a class of software that's developed to manage a distributed system. It's responsible for implementing the complex details of a distributed system, and providing a 4
• 5.
  virtual development platformof a single system with component-based model, that's actually distributed over many computers. You're (as the application developer) responsible only for developing software-components and deploy them on the middleware, which will run them in managed environment of a distributed computing model, providing them was all needed system services. 3. Java Enterprise Edition Platform The Java EE is complete platform for developing multi-tiered distributed applications. It consists of: • The Java Virtual Machine (JVM). • The Java SE APIs (SE APIs). • The Java EE Application Server. • The Java EE APIs (EE APIs). As you must have noticed, the Java EE Platform consists of the Java SE Platform (JVM and SE APIs), with two additions: - The Java EE Application Server A middleware that supports the distributed system. It implements the system programming part of the distributed system, and provides a platform for developing component-based applications that run in a managed environment. - The Java EE APIs 5
• 6.
  A set ofAPIs that provides the common functionalities needed in enterprise systems. 3.1. Application Architecture In Java EE, applications are divided into four teirs: Enterprise Information System (EIS) tier, business tier, web tier, and client tier. 3.1.1. EIS Tier The Enterprise information system tier contains the enterprise information system server that contains the data of the application. It is usually the Relational DBMS. 3.1.2 Business Tier The Business tier contains business components where business logic is 2 implemented. 6
• 7.
  3.1.3 Web Tier 3 The web tier contains web components that generate dynamic web pages over hyper text transfer protocol (HTTP). Pages are accessed using the web browser, that's known in Java EE as the thin client. 3.1.4 Client Tier The client tier access system services. There are two variations of the client tier: 1. Swing Application: in case the client access the business components. 2. Web Browser: in case the client access the web components. 3.2. Application Design In Java EE, applications are developed using a component-based model. The Java EE platform provides different types of components that can be developed and deployed on the application server. We will study two different types of components: business components and web components. 3.2.1 Business Components Business components (known as: Enterprise JavaBeans, or EJB in short) are developed to implement business logic. EJBs are standard components that can be 4 accessed remotely using RMI . There're two types of EJBs: - Session Beans. - Message-Driven Beans. 3.2.2 Web Components Web components are responsible for generating web pages and documents. There're two types of Web components: - Servlets. - JSPs (JavaServer Pages). 2. Business logic includes processing data for operations like accounting, inventory, HR management, and so on. 3. Web pages are usually HTML pages that are either static or dynamic. In case of static web pages, content does not change and every time the browser visits the page it receives the same content. Dynamic pages however, are created dynamically when the browser request the page. The page is generated dynamically and the browser may receives different content each time it visits the page. 4. Remote Method Invocation: a Java API for implementing remote procedure calls (RPCs). 7
• 8.
  5 There are significantdifferences between Plain Old Java Objects (POJOS) and Java EE components: - Components are standardized. Java EE components should implement certain interfaces, extend some class, or following any other standard required by the specification to be certified as a component in Java EE platform. Unlike POJOs, classes should not follow any certain standard requirement. - Components are managed by the Java EE Application Server. Enterprise applications do not run by themselves as standalone applications, instead, they should run in a managed environment provided by the Java EE application server. Thus, components are developed and compiled, then deployed on the Java EE application server. Application server then runs them in its managed environment. Unlike POJOs, where we used to run the class with the main() method. In other words, Java EE applications does not have main() method. In addition, components are not accessed directly. The client tier access components using standard interfaces. Unlike POJOs, where we used to access the object directly using its reference. In other words, the application server acts as a service layer between the client tier and other tiers. - Components are not instantiated using new operator. In distributed systems, components are not accessed directly. Instead, component are named and located. When the developer deploys a component on the application server, she gives the component a unique name. The client then asks the application server to locate the component, and instantiate it for him, using its name. In other words, the client does not instantiate the component, but the application server is responsible for instantiating and returning a reference to the component for the sake of the client. - All components are developed for the sake of the client. Take in concentration that the ultimate purpose of Java EE is to develop distributed multi-tiered applications. Thus, all components are developed and deployed to provide a service that's actually used by a consumer, the client. Every EE application we are going to develop will has clients to access our developed system. Either we will develop the client ourselves in case of Swing Application clients, or the client will be the web browser in case of accessing web components. 5. The ordinary java objects we used to develop and use in Java SE applications. 8
• 9.
  4. The JavaEE Application Server The Java EE application server is a specification by Sun Microsystems that describes a middleware system. The system should provide a managed environment for Java EE components, in addition to an implementation to the Java EE APIs. Java EE APIs are a specification, as well. Any system that implements the Java EE application server and Java EE APIs specification is called Java EE application server, or application server in short. In fact, any vendor can implement the specification, but should get certified first by Sun that her implemntation is a valid application server. There are dozens of application servers in the market. Here is a list of some application servers and their vendors: Application Server Vendor Type Glassfish Sun Open Source/Free JBoss RedHat Open Source/Free Edition WebSphere IBM Open Source/Free Edition OC4J ORACLE Commercial We will use JBoss application server, that can be downloaded from: http://www.jboss.org/. Your role, as the application developer, is to only develop Java EE components according to the specification, then deploy them on the application server. Since both the components and application servers are developed according to the specification, any Java EE application can run on any Java EE application server, which is the sexiest property of the Java EE platform. 4.1. Runtime The application server provides the runtime platform in the form of containers. A container is a component of the application server that manages a specific type of components and provide them with all the needed services and Java EE API implementaiton. We will study two types of containers: - Business Container The business container is responsible for managing business components and providing them with runtime services like: RMI communication, database manipulation, transaction services, etc. - Web Container The web container is responsible for managing web components and providing them with runtime services like: HTTP communication, database manipulation, etc. 9
• 10.
  4.2. Services The applicationserver provides each container with a set of required services. The service is implemented and supported to the application developer via Java EE APIs. Services include: - Remote Method Invocation (RMI). - HTTP (Hypertext transfer protocol). - HTTPs (HTTP over SSL). - Java Messaging Service (JMS). Supported via JMS APIs. - Java Naming and Directory Interface. Supported via JNDI APIs. - Distributed Transactions Management. Supported via Java Transactions APIs (JTA). - Java Database Connectivity. Supported via JDBC APIs. - Persistence. Supported via Java Persistence API (JPA). - Java Authentication and Authorization Service (JAAS). 4.3. Final Image - Java EE Architecture 10