OCEJBCD -10. Implementing transactions

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

10.1. Describe transaction demarcation management

JTA (Java Transaction API).

Transaction demarcation involves indicating when a transaction begin and ends.

There are three kind of demarcation management:

• Bean managed
• Container managed (by default)
• Client managed

10.2. Implement CMT

CMT (Container Managed transaction)

Transaction attributes

Transactions attributes can be used at class or method level.

• Required

@TransactionAttribute(REQUIRED)
Start transaction or continue with client transaction.

• RequiresNew

@TransactionAttribute(REQUIRES_NEW)
Start new transaction (stop client and resume later).

• Mandatory

@TransactionAttribute(MANDATORY)
Always with client transaction, if not throws an exception

• Not supported

@TransactionAttribute(NOT_SUPPORTED)
No transaction (if client running within trans stop and resume later)

• Supports

@TransactionAttribute(SUPPORTS)
Bean runs within client transaction if related only.

• Never

@TransactionAttribute(NEVER)

10.3. Interact programmatically with an ongoing CMT transaction

A transaction will rollback automatically in CMT when a system exception ocurrs. It can be invoked by setRollbackOnly() method.

Methods not allowed in CMT are ( commit, rollback, getUserTransaction...)

10.4. Implement BMT Apply transactions to messaging

@TransactionManagement(BEAN)

Use of commit, rollback, getUserTransaction(setTransactionTimeout..etc)

Not allowed get/setRollbackOnly()

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

OCEJBCD -9.Implementing Interceptor Classes and Methods

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

9.1. Describe interceptors and interceptor classes

An interceptor is a class used to intercept calls to EJB.

They are usually used for logging, authentication or validation purposes.

A typical example can be logging at the beginning and end of a method. This can be a tedious task, imagine to create logging entries for a big application. With interceptors it is avoided this duplicity of code, only specifying this behaviour in an interceptor class and used by annotations in the classes of your application.

The interceptor can be used by all methods of the class if we annotate our bean with :
@Interceptors(MyInterceptor.class)
We can apply more than one interceptor separating list with comma:
@Interceptors(MyInterceptor1.class, MyInterceptor2.class)

Or we can define which methods will be intercepted with annotations at method level with: 
@Interceptors(MyInterceptor.class)

Also we can exclude class interceptors to be executed in some methods with:
@ExcludeClassInterceptors

9.2. Create a business interceptor method in the enterprise bean class

Also we can define an interceptor method inside the bean class. 
This interceptor method will be the last one to be invoked before the actual bean method:



2016-08-28T10:50:05.833+0100|Info: Entering method:public void example.TestBean.firstMethod()
2016-08-28T10:50:05.833+0100|Info: Leaving method:public void example.TestBean.firstMethod()

9.3. Create an interceptor class

There are three kinds of interceptor method:

• method interceptor methods for methods of a target class.
• @AroundInvoke
• timeout method interceptor methods for timeout methods of a target class 
• @AroundTimeout
• lifecycle callback interceptor methods : 
• @PostConstruct and @PreDestroy 
• (and for @PrePassivateand @PostActivate methods of EJB stateful session beans).

9.5. Associate multiple business interceptor methods with an enterprise bean

Applying the interceptor class defined in the previous section to our test bean. In the output log we can verify that two interceptors method are being invoked for firstMethod: first the interceptor at class level, and second the one at method level.
The second method will be called only by the interceptor at bean method due to annotation of excluding class interceptors.




2016-08-28T11:07:55.689+0100|Info: **Begin public void example.TestBean.firstMethod()
2016-08-28T11:07:55.689+0100|Info: Entering method:public void example.TestBean.firstMethod()
2016-08-28T11:07:55.689+0100|Info: Leaving method:public void example.TestBean.firstMethod()
2016-08-28T11:07:55.689+0100|Info: **End public void example.TestBean.firstMethod()
2016-08-28T11:07:55.689+0100|Info: Entering method:public void example.TestBean.secondMethod()

2016-08-28T11:07:55.689+0100|Info: Leaving method:public void example.TestBean.secondMethod()

9.6. Include life cycle callback interceptor methods in an interceptor class

The lifecycle of an interceptor instance is the same as that of the target class instance with which it is associated.

@PostConstruct @PreDestroy
public void interceptLifeCycle(InvocationContext ctx){...

This callbacks can be used to initialize the state of the bean.
Because callbacks have not return value, the return type must be void.

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

RMI and Sockets

This post will compare the solution to choose when a program from a client calls a program in the server in distributed applications. Both scenarios can be implemented using Sockets or RMI. But when this question arises what are the advantages and inconveniences of every one? 

RMI vs Sockets

• Solution:
• Sockets allow to build a custom solution
• RMI is a prebuilt architecture. 

• Language:
• RMI provides remote communication between programs written in Java.
• Sockets allow communication with any other language different from Java.

• Resources
• RMI consumes more resources in the network.

• Complexity
• RMI hides serialization and network communication details, also thread management is done by RMI system. RMI also allows dynamic class loading.

RMI Architecture

Fork Join Framework

Fork Join technology was introduced with Java version 7, to be able to deal with parallel concurrency.

This technology is based in splitting a task recursively until it is small enough to be performed. The sub-tasks will be executed in parallel in different threads. When the sub-tasks have finished, then a join is invoked to merge the results into one.

Example. The following example is very simple. It consists on a function that calculates consecutive natural numbers in an interval. The class will receive two parameters (number min. and max.). If this interval is greater than 6 (in the example) the task will be split into two sub-tasks.

The class extends RecursiveTask<Long> and implements method compute(). As our class is a task it will be invoked by ForkJoinPool



Output console:
Splitting task
Computing
Result1: 21
Computing
Result1: 57
Result2: 78

OCEJBCD -8.Using Timer Services

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

8.1. Describe timer services

Timer services is a service of scheduling notifications in the bean container.
(Available for all bean types except from stateful session beans).

Timers services can be created 
-Programmatically, using TimerService interface
-By annotations @Schedule

Functionality that a bean with timer services can perform:

• Create a timer notification callback
• Process a timer notification callback
• Manage timer objects (list notification, cancel timer etc...)

The timers can be classified as:

• Single notifications
• Interval (multiple notifications at specified interval)
• Calendar based schedule 
• second, minute, hour 
• dayOfWeek, dayOfMonth
• month
• year
• Expressions : wildcard (*), list(,), range(-),intervals(/)

8.2. Create a timer notification callback

• By timerService interface
• Single:

          Timer timer = timerService.createSingleActionTimer(data, new             TimerConfig());

• Calendar based: using javax.ejb.ScheduleExpression

          ScheduleExpression schedule = new ScheduleExpression();
          schedule.dayOfMonth("1,Last");//first and last day of the month
          Timer timer = timerService.createCalendarTimer(schedule);

       When a timer expires the container calls the bean's method annotated              with @Timeout (*Note: this method must return void and the argument is optional).

• By annotation

       Example : @Schedule(dayOfWeek="Mon", hour="0")

8.3. Process a timer notification callback 

Example with timerService interface

To execute this bean we need to make a call from a client (for example a servlet class) to the method setTimer that indicates the milliseconds after the timeout will happen: scheduleBean.setTimer(30000);

Results in server console:
2016-03-26T11:58:33.313+0000|Info: TestBean was successfully deployed in 362 milliseconds.
2016-03-26T12:00:26.816+0000|Info: the next timeout is programed Sat Mar 26 12:00:56 GMT 2016 happening in 29968 milliseconds

2016-03-26T12:00:56.787+0000|Info: Timeout occurred

Example with schedule annotation
The following example is a stateless session bean, with a timer scheduled every 1 min.
When it is deployed in the server, after 1 minute the timeout method is called:

Results in server console:
2016-03-26T10:03:00.001+0000|Info: Automatic timeout occured
2016-03-26T10:04:00.002+0000|Info: Automatic timeout occured
2016-03-26T10:05:00.001+0000|Info: Automatic timeout occured
2016-03-26T10:06:00.002+0000|Info: Automatic timeout occured
2016-03-26T10:07:00.002+0000|Info: Automatic timeout occured

8.4. Manage timer objects

-Cancel() - cancel a timer
-getHandle() - obtain a serializable object
-getTimeRemaining()
-getNextTimeout()
-getInfo()

When using container-managed transactions @Timeout can use attributes: Required or RequiredNew to preserve transaction integrity, which means that in case of rollback the timeout will be reset.

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

How to upload files using Servlet technology

This time, the aim of this post is to describe how to upload files to a server using servlet technology.

Upload files with Java technology is easy, we only need to create a JSP page and a servlet class, but we have to take into account this form will be a bit different than usual:

• Method POST 
• multipart/form-data 

Form declaration in the JSP:

<form action="upload" method="post" enctype="multipart/form-data">

Annotations in the servlet declaration:

@WebServlet("/UploadServlet")

@MultipartConfig

public class UploadServlet extends HttpServlet {

JSP code:

Servlet code:

* This example project is using servlet library version 3.0 or above.

Method post - summary:

1) Getting the file attribute from the request object. file attribute corresponds
with the name of the input type ="file" in the JSP. The object obtained is Part type.

2) Obtain the name of the file uploaded:
using filePart.getHeader to obtain the value of the property filename.

3) Defining an object OutputStream with the destination path, plus the file name).

4) Read the content of the object filepart obtained in point 1, and writing the bytes into the ouputStream (in the destination).


Running the application:  
http://localhost:8080/UploadWeb/

 Select a file from the local disk and click the button to submit the request:

The server logs will display this message:

INFO: Reloading Context with name [/UploadWeb] is completed

Checking the execution result (the path selected in the example is also a local directory C:/uploads), we can verify the file has been uploaded successfully to the selected destination.

OCEJBCD -7.Developing Message-Driven Beans

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification

7.1. Understand the short-comings of using session beans as messaging consumers

Using a session bean as consumer has the problem of blocking server resources, because it will be a synchronous process (instead of asynchronous processing with message-driven bean ).

Using session bean receive() method will block the thread indefinitely until the message becomes available :

Message m = consumer.receive()

How to handle the blocking synchronous problem with session bean? Using a timed synchronous receive.

A parameter can be used to indicate the time waiting the message to arrive:
Message m = consumer.receive(1); //time in milliseconds
Message m = consumer.receiveNoWait();

See an example of Writing the Clients for the Synchronous Receive.

7.2. Describe the properties and life cycle of message-driven beans

7.3. Create a JMS message-driven bean

To create a message-driven bean the bean class:

• is annotated with @MessageDriven
• @ActivationConfigProperty Settings(*):
• Acknownledge_Mode
• Auto-Acknownledge
• Client-Acknownledge
• Dups-ok-aknownledge
• DestinationType
• Queue
• Topic
• Subscription durability
• Durable
• Non-durable
• ClientId
• SusbscriptionName
• MessageSelector 
• AddressList
• implements MessageListener inteface
• overrides onMessage() method

(*)Example of @ActivationConfigProperty configuration:

@MessageDriven(mappedName = "jms/Topic", activationConfig =  {
    @ActivationConfigProperty(propertyName = "messageSelector",
            propertyValue = "NewsType = 'Sports' ")
    , @ActivationConfigProperty(propertyName = "subscriptionDurability",
            propertyValue = "Durable")
    , @ActivationConfigProperty(propertyName = "clientId",
            propertyValue = "MyID")
    , @ActivationConfigProperty(propertyName = "subscriptionName",
            propertyValue = "MySub")
    })

There is an example of a message-driven bean class at the last section of this post.

7.4. Create life cycle event handlers for a JMS message-driven bean

See this post : Configuring a JMS environment, to know how to configure a server and include jms resources as connection factory and destination

Once the connection factory and the destination(Queue or topic) have been created, we can use dependency injection to use them in our message beans

like:

• Connection factory

@Resource(name="jms/MyConnectionFactory")
private QueueConnectionFactory connectionFactory;

• Destination

@Resource(name="jms/MyQueue")

private Queue queue;

With the connectionFactory we can define the context to create the producer and the consumer.

• Connection and Session

Connection con = connectionFactory.createConnection();
Session session = con.createSession(true,0);

• MessageProducer

MessageProducer producer = session.createProducer(topic);

• Message types(TextMessage, MapMessage, ObjectMessage and others).

TextMessage message = session.createTextMessage();


7.5. Configure a JMS message-driven bean

Here is an example of using a message-driven bean:

• A client where the user request to send a message (with a jsp page and a servlet, in the same way as examples described in previous posts : calling a session bean from a servlet).
• A session bean (stateless) that will define a producer that put the message in the JMS queue.
• A message-driven bean that will consume the message from the queue.

Session bean:

Message bean:

When executing the application in our configured server:

Write a message in the textarea and submit the message.
In the log can be read the producer sending the message and the consumer receiving it:
producer sends the message...
consumer receives message...

OCEJBCD (SCBCD) - 1Z0-895 - Enterprise JavaBeans Developer Certification