EJB
EJB is a server-side component model sun, best use is to deploy distributed applications, of course, there are many ways to implement distributed applications, like Microsoft's. Net technology. With the advantages of cross-platform java, using EJB technology deployment of distributed systems can be limited to a specific platform. EJB (Enterprise JavaBean) is part of J2EE defines a component-based enterprise for the development of multi-standard applications. Features include Web services support and core development tools (SDK). In the J2EE, Enterprise Java Beans (EJB) called the Java Enterprise Bean, is Java's core code, namely the session Bean (Session Bean), Entity Bean (Entity Bean) and Message-Driven Bean (MessageDriven Bean).
directory
EJB profile one. What is EJB? II. Sun expectations of EJB III. Why EJB? IV. EJB architecture five. Simplified introduction to start programming model background EJB3.0 Overview of EJB persistence model of new comments stateless session bean stateful session bean message-driven Bean Entity Bean Entity associated with O / R mapping model of the client program where clause of EJB QL supports subqueries nested diversity copied to the clipboard Print View EJB specification document processing mode associated with the other concluding remarks on the EJB3.1 About 1.Session Bean EJB for business logic, it can be a state, it can be stateless. Whenever a client request,
cheap paul smith, the container will select a Session
EJB
Bean to the client service. Session Bean can directly access the database, but more often, it will access data through the Entity Bean. 2.Entity Bean is the domain model objects, used to implement O / R mapping, is responsible for mapping tables in the database records in the Entity objects in memory, in fact, create a new Entity Bean object representing a record, delete an Entity Bean will also delete the corresponding records from the database, modify an Entity Bean, the container will automatically Entity Bean state and database synchronization. EJB2.0 3.MessageDriven Bean is introduced in the new enterprise Bean, which is based on JMS messages, only receive JMS messages sent by the client and processing. MDB is actually an asynchronous stateless Session Bean, MDB after the client calls without having to wait, return immediately, MDB asynchronous handling of customer requests. This need for asynchronous processing requests for the occasion, such as order processing, so that clients can avoid a long wait for a method call until it returns the results. SUN's J2EE EJB is actually the set of norms, and provides a series of API used to implement the EJB EJB concepts into products. EJB is BEANS, BEANS What is the concept, that is, must have a capacity of her, let her Jin Tang can be made is the place must have a container. EJB must survive in the EJB container. the container, but extremely powerful! her first package you BEAN, EJB client program and you never actually write the EJB directly deal between them through the HOME / REMOTE interface to a relationship. It is responsible for all of your eating and drinking Lazard BEAN sleep, such as BEAN's persistence, security, transaction management ... one. What is EJB? a Technical specifications: EJB Technically not a Transactions (Transactional) data storage (Persistent) Security (Secure) II. Sun expectations of EJB provides a standard distribution, based on the OO component architecture shield the functional requirements of complex system-level Write once, run anywhere and non-Java interoperability between applications compatible with CORBA standard three. Why EJB? EJB server to complete the ) to support multiple business operations while the transaction succeed, or all of the failures can be described by code defined outside the transaction-level
EJB
scalability of EJB applications according to the growth of your EJB server is often also extended provides load balancing and security: the provision of resources by the EJB server access control four. EJB architecture framework in order to meet the goal, the specification describes the server (Server) container (Container) Class (Class) and instances (Instance) Home and Remote interface to the client (Client) five. simplified programming model focuses on the business logic implementation: EJB is responsible for life-cycle (lifecycle), data storage (persistence), transaction semantics (transactional semantic), safety (security), ... GM programming model: high-level services is the programming language API Java EJB (business logic) that a specific business areas (such as banks, retail and other industries) to adapt the logic. It consists of running the enterprise bean business logic processing. An enterprise bean from the client to accept data, process it and send it to the enterprise information system tier for storage; same time it can get data from memory, processing and sends to the client application. There are three types of enterprise beans: session beans, entity beans and message-driven beans. Session bean with the client described a brief conversation. When the client after the execution, session bean and its data will disappear; is a corresponding entity bean is stored in the database table describing a row in the persistent data, or services if the client terminates the end, the underlying Services will be responsible for the entity bean data storage. Message-driven bean with a session bean and Java message service (JMS) message listener function, which allows a business component to receive JMS messages asynchronously. EJB2.0 to EJB3.0 introduction to the complexity, in the Spring and Hibernate and other lightweight frame appears, a large number of users turned to the voice of the people,
EJB
EJB3 long-awaited .0 specification finally released. In this article, the new specifications will be an overview of the presentation, including new metadata support, EJBQL changes, the entity bean Bean model to access the context of the new method and runtime environment and so on. The authors also discuss the EJB in the future to make the adjustment and development of EJB3.0 specifications and other relationships. Start in any case the complexity of the EJB in the J2EE architecture to make the performance have not been good. J2EE EJB architecture is probably the only one that did not honor its simple components to develop and increase productivity. EJB3.0 specification are trying to make efforts in this regard to reduce the complexity of its development. EJB3.0 developers to reduce the workload of the underlying development, it is to cancel or minimize a lot of (previously these are to be achieved) callback method implementation, and reduces the entity Bean and O / R mapping model complexity. In this article, I will first introduce several major changes in EJB3.0. It further in-depth understanding of EJB3.0 is very important. Then, I will describe a higher level have been submitted to the EJB3.0 specification details, and one of the new specification to explain the change: entities Bean, O / R mapping model, entity relationship model, and EJB QL (EJB Query Language) and so on. EJB3.0 in the background are the two important changes: The program uses the annotation tool Java5 and based on Hibernate's O / R mapping model.
EJB
Java5 of metadata tools Java5 (formerly known as J2SE1.5 or Tiger) to add a new program annotation tools. With this tool you can customize the comment tags, custom tags to these comments fields, methods, classes and more. These comments will not affect the process of semantics, but can be tools (compile time or run time) to interpret these tags and generate additional content (such as the deployment descriptor), or to force some of the required run-time behavior (such as EJB group pieces of state property). Note the resolution can resolve the source file (such as compiler or the IDE tool) or use the APIs Java5 reflection. Comments can only be defined in the source code level. Since all the draft is submitted to the EJB3.0 has a comment tag in the run-time RetentionPolicy, it will increase the class file storage space, but it gave the container manufacturers and tool manufacturers to bring the convenience. Hibernate Hibernate is very popular at present, it is the development of the source code Java O / R mapping framework, which aims to developers from the tedious data persistence programming freed. It also has a standard HQL (Hibernate Query Language) language, you can see the new EJB QL in its shadow. Hibernate in processing such as data queries, updates, connection pooling, transaction handling, processing, entity-relationship is very simple. Overview of the EJB3.0 specification has been submitted to the major changes involve two areas: 1. A set of annotation-based EJB programming model, coupled with the adoption of defined EJB2.1 deployment descriptors and several interface definitions application behavior. 2. New entity Bean persistence model, EJBQL there are many important changes. Some of the above proposals, such as: a new client-side programming model, the use of business interface, and the life cycle of an entity Bean. Please note that EJB2.1 programming model (including the deployment descriptors and home / remote interfaces) is still valid. The new simplified model does not completely replace the EJB2.1 model. EJB specification EJB annotation is an important organizational goal is to reduce the number of the original code, and they do give a perfect and simple way. In
EJB
EJB3.0 of, any type of enterprise Bean is just a simple addition of the appropriate Java object annotation (POJO). Notes can be used to define the bean's business interface, O / R mapping information, resource referral information, the effect is defined as in EJB2.1 deployment descriptors and interfaces are the same. In EJB3.0 deployment descriptors are no longer in need of a; home interface, and no, you do not have to implement the business interface (the container for you to complete these things.) For example, you can use the @ Stateless annotation marks the Java class declaration class as a stateless session bean. For stateful session bean is, @ Remove annotation can be used to tag a particular way, through this comment that this method is called after the bean instance will be removed. To reduce the description of the component description information, standardized organization also adopted by the abnormal configuration (configuration-by-exception) means, meaning that all the comments you can provide an explicit default value, so the majority of general information can inferred accordingly. The new persistence model is a new entity bean annotations added a simple Java object (POJO). Once it has been EntityManager to access it becomes a persistent object, and became the persistence context (context) part. A persistence context with a transaction context is loosely coupled; Strictly speaking, it implies co-exist with a transaction session. Entity relationship is defined through annotations, O / R mapping is, and provides several different databases and standards, in EJB2.1 of these
EJB
by the developer to design their own patterns or other technology to complete (for example, since the primary key growth strategy.) Now is the time-depth research to learn more about the draft of the EJB3.0. Let us begin to explore all of four enterprise-class EJB bean, and see them in the new specification is like. EJB3.0 stateless session bean in the specification, write a stateless session bean (SLSB) requires only a simple Java file and add the class level @ Stateless annotation to it. The bean can extend javax.ejb.SessionBean interface, but these are not necessary. SLSB no longer need a home interface, there is no type of EJB longer need it. Bean class can implement the business interface can not implement it. If you do not implement any business interface, the business interface will be generated by any public way. If only a few business method will be exposed to the business interface, these methods can be used @ BusinessMethod comments. By default, all generated interfaces are local (local) interfaces, you can also use the @ Remote annotation to declare the interface for the remote (remote) interface. The following few lines of code to define a HelloWorldbean the. In EJB2.1 bean in the same need at least two interfaces, an implementation class and a few empty implementation methods, coupled with the deployment descriptor. import javax.ejb .*; / ** * A stateless session bean requesting that a remote business * interface be generated for it. * /; @ Stateless; @ Remote public class HelloWorldBean {public String sayHello () {return !!!the beginning of their own (in EJB2.1 is by ejbCreate () to achieve.) EJB3.0 specification in the initialization proposed in these methods can be customized to complete, and send them exposed to the business interface. Before using this bean from the client to call the appropriate initialization method. Specification as to whether the current organization to provide a comment to mark a method used to initialize the still controversial. Bean provider can use the @ Remove annotation to mark any SFSB way to illustrate this method is called after the bean instance will be removed. Similarly, the specification is still under discussion whether the organization should have a mechanism to handle this special case, that is, when this method is the case of abnormal bean instance is removed. Here is my personal point of view these questions: Should there be a note to indicate a method to initialize? My view is that there should be, so the container can call other methods before calling a method for at least the early initialization. This not only avoids unnecessary errors (due to not call the initialization method) and can more clearly determine whether the container can be reused SFSB instance. I am going to put off this issue, considered as a standardized method of organization only to provide a comment to state that it is an initialization method. For my point of view the second question is yes. This is conducive to co-Bean provider client program to control it. Only one question left: that is, once the call that fails, it can remove the bean instance? Answer is no, but it will be removed when the session ends. Message-Driven Bean Message Driven Bean is the only must implement a business interface Bean. The interface that supports what kind of bean messaging system. For JMS-based MDB, this interface is javax.jms.MessageListener. Note MDB business interface is not a real business interface, it is just a message interface. Entity Bean Entity Bean using the @ Entity annotation to mark all the attributes in the entity bean / field do not have to use the @ Transient annotation to mark. Entity bean's persistent fields through JavaBean-style mechanism or declared as public / protected fields to achieve. Entity bean can be used to describe the status of helper classes, but instances of these classes are not persistent unique (persistent identity) characteristics (ie, uniquely identifies the bean field, etc.), in fact, these helper classes with them entity bean instances are closely linked; and sharing of these objects is still way to access a non-physical objects. Entity Bean association between EJB3.0 supports two-way one-way relationship of cooperation, they can be one to one, many, many to one or many to many association. However, the two ends of a bidirectional association also divided into its own end (owning side) and the other side (inverse side) different terminal. Notice to the database side is responsible for its associated changes. For many to many association must be explicitly declared its end. In fact the other side were comments by isInverse = true (which itself illustrates the client will not have to be inferred but by another section.) It seems the above description, specification organization also said that a simple change to the EJB you? O / R mapping EJB3.0 in the O / R mapping model has also been important changes, it is from the original abstract-persistence-schema-based become now
EJB
in the Hibernate-inspired patterns. Despite the current norm organization also discuss this model, but there will be a clear in the next version of the draft. For example, O / R mapping model bean class through annotations to declare. But also pointed out that this method corresponds to the specific tables and fields. O / R mapping model provides a set of its own SQL; and SQL in addition to providing some basic things, to support the development of certain high-level functions. For example, there is a statement through the @ Column annotation field columnDefinition, you can write this SQL: columnDefinition = get a bean's business interface reference. You can also use another annotation @ javax.ejb.EJBContext.lookup () to perform the above operation, but the specification does not tell us a regular Java Bean client how to obtain an instance, because the client is an ordinary Java run in a client container, it can not access the @ javax.ejb.EJBContex object. Now there is another mechanism to accomplish the above work is to use a super-context object: @ javax.ejb.Context (). But the specification does not indicate how the client use this object. EJB QL EJB QL @ NamedQuery to comment through. The comments are two members of the property name and the queryString. Once you have defined these properties, you can access EntityManager.createNamedQuery (name) to point to the query. You can also create a standard JDBC-style query and use EntityManager.createQuery (ejbqlString) or EntityManager.createNativeQuery (nativeSqlString) (This method is used to execute a local query) to execute the query. EJB QL has two places to define its parameters. javax.ejb.Query interface provides the definition of parameters, point to query and update data, etc. methods. Here is an example of a query EJBQL point: View the print copy to the clipboard. ..; @ NamedQuery (name = .; @ Inject public EntityManager em; customers = em.createNamedQuery (updates and deletes. Directly support the Inner and outer connections. FETCH JOIN run your associated entities that, Order can specify only a field check. Query can return more than one result value. In fact, you can return to a dependent classes such as follows: SELECT new CustomerDetails (c id, c.status, o.count) FROM Customer c JOIN c.orders o WHERE o.count> 100 l support group by and having . Support for nested sub-queries where clause in the draft submitted by the EJB3.0, EJB QL is very close to the standard SQL. In fact the specification or even directly support local SQL (as we did above.) That some programmers may be some not very clear, we will explain below in more detail. Diversity and license (Method permissions) and via @ MethodPermissions @ Unchecked annotation to declare; Similarly, the transaction attribute can also declare @ TransactionAttribute comment. Retain the reference to the specification of resources and resource environment references. These annotations can be declared as, but there are some subtle differences. For example, the context (context) environment must be controlled by injecting tool. Container under the bean reference to the external environment has been automatically initialized an appropriate instance variable declaration. For example, you may like this to get a data source (DataSource):; @ Resource (name = (name) then customerDB which is considered the default. When all references to property can be time @ Injec comments can be written: View the print copy to the clipboard 1.; @ Inject public DataSource customerDB;; @ Inject public DataSource customerDB; container is responsible for the data at run-time initialization customerDB source instance. Prior to this deployment of personnel to be defined in the container properties of these resources. Even better news is: those who previously would have gone for detection of abnormalities. You can declare an arbitrary application exceptions, without having to throw it or catch CreateException and FinderException such other similar exceptions. Encapsulated in the container will throw javax.ejb.EJBException or in the system-level exception IllegalArgumentException thrown only when necessary or IllegalStateException. EJB file processing mode we conclude this section, let me quickly look at the EJB container provider, possible changes in processing mode. Specification is no clear position on this, but I can think of at least two modes. One approach is to first use EJB file to generate a file similar to EJB2.1 deployment model (including the necessary interfaces and deployment descriptors) and then use the similar way to deploy the EJB2.1 EJB components. Of course, this deployment descriptor generated may not be standard but it can be solved with a container for EJB2.1 and EJB3.0 compatibility issues. The picture below describes this process. Another method is to put a similar deployment of JSP care model. You can put an EJB file into a predefined directory, and then the EJB container will identify and deal with it, and then deployed and made available. This method can be built on top of that method above, repeated deployments in support of a great help. Considering the simplicity of deployment is one of the purposes EJB3.0 specification, I sincerely hope that the next draft to determine when a model came out (at least to have an informal). What are your thoughts? EJB3.0 specification for the development is orderly, in order to make easier the development of EJB, EJB specification organization's efforts are obvious to all. Like they say, everything will become simple, but to do this is not easy. 50 currently defined comment tags (as well as several will be released the next draft), each with its own default rules and other operations. Of course, I really do not want a replica of EJB3.0 into EJB2.1 Finally, I can not help but mention some of my own point of view: first, the specification does become easier to deploy repeatedly, and there is a simple model to access the runtime environment. I still think the home interface should be abandoned. In the early EJB specification, an entity bean is used to map a persistent storage. In theory (and perhaps only in theory) may require the entity bean mapped to a legacy of the EIS (enterprise information system) system. Consideration for future expansion is beneficial to do so, and can make more use of business data model entity bean. Thus making the complexity of its associated entity bean is not good. Presented in this draft, an entity bean is just a database mapping. And is based on non-abstract persistence model and simple data access patterns easier development. I have reservations about the model changes, I think that is included in the EJB SQL script fragment is not a good idea. Totally opposed to some developers include some of the My view is that these SQLness is good, pursuant to which we can clearly know that we in the end to the database for something. However, it seems to me some SQL section is not very good, such as columnDefinition = Despite the support of the local SQL appears to be very attractive, in fact, embedded SQL code in the EJB is a very bad idea. Of course, some way to avoid hard-coded in the EJB in SQL, but it should be described in the specification, but can not be defined in some developer's own model. Suppose @ Table annotation is only used for classes. At run time by @ Table annotation name attribute defines the name of the table must correspond to an actual database table. Specification should be given clear instructions to this and consistent pattern. Specification also need a clearer description of client-side programming model, especially the ordinary java client. All specifications are assumed or implied reference to the use of EJB client. And the specification of the backward compatibility aspects of the client did not give a clear view also. Transient annotation should be renamed to avoid conflict with existing transient keyword. In fact, at this point we are more willing to look a little departure from the principles of configuration-by-exception and the @ Persistent annotation to define a clear definition of persistent fields @ Persistent annotation can simply be a marker annotation or it can have several attributes to associate O / R mapping annotations. Currently associated with the other standards that may affect the JSR EJB3.0 has JSR175 (java language meta-data tool) and JSR181 (Java Web Services Metadata) JSR175 has been initially completed and will not have much conflict EJB3.0; but JSR181 and there are two related areas EJB3.0: Web service interfaces: EJB specification will introduce a mechanism so that it can adapt to JSR181 to a bean is implemented as a Web service and tell the client how the Web service call. JSR 181 plans to use different mechanisms to deal with security issues. In the early EJB specification recommends using a consistent mechanism (MethodPermissions), but JSR 181 plans to use a slightly different way (SecurityRoles and SecurityIdentity comments). RunAs annotation the same definition of these promised differences. This issue is still resolved in the final layer in the J2EE specification to maintain its consistency. In the J2EE 1.5 specification in some development may be associated with EJB3.0. In addition to the above mentioned no other than a few associated with the development of norms in conflict with EJB3.0. About EJB3.1 for EJB, all the reconstruction is the priority. Now the cup has been emptied, which is very beneficial for us, but also a bold move forward without the burden. EJB3.1 once again introduced a series of new features, tend to tap the potential of technology. In my view, EJB3.1 is definitely an important release that will feature those who desire to bring the long-term development of people in front, more able to meet the latest enterprise application development, while the adoption of EJB will do it again a huge contribution. Currently, EJB 3.1 final release version has been released. EJB3.1 architecture model structure relative to the old mode as some of the extensions and for a number of API simplification: Simplified model for the local to the interface in addition to the local business can achieve the same session bean · war file directly using an ejb component package do not need an ejb-jar · embedded in the Java SE environment, the implementation of EJB · API provides a simple introduction to share and support consistency of application data access Singleton session bean EJB component is automatically created based on the EJB Timer timer expression schedule calls for an asynchronous session bean EJB defined set of light quantum, the subset of Java EE can provide some documentation (EJB lite) · ejb components used to query a Global JNDInames (uniform global JNDI names) Rules Conclusion In the EJB development easy and efficient, we still have a long way to go. EJB specification development organizations, the difficulty in reducing the play to a great start. O / R mapping model is still in the early stages of the proposal, standard organizations are improving it. I hope that it will not be too complicated and not with the SQL over the coupling. Atlas Atlas entry for more extended reading: 1
Advanced Technology in the Science Wing Alliance Training Center
2
http://hi.baidu.com/skychen1900
3
EJB3.0
http://ejb.java2000.net/ JAVA century information networks
4
http://java.sun.com/javaee/technologies/Enterprise JavaBeans 3.1
Open Category: JAVA, J2EE me to perfect the
开通空间啦
Linear Mode
