EJBs only for session facade?

Hi all,
With EJB 2.1 many were saying that it is good to place POJOs behind a session facade of ejbs for a distributed application.
This will allow a developer to test his business logic outside the container.
But with EJB 3.0, all ejbs are pojos, so can i use ejbs for all my business logic?
Is it advisable?
Will this have any performance problem?
Kindly suggest.

Hi,
Thanks for your reply.
I have a distributed application with numerous web clients as well as a few multithreaded application clients accessing the business logic layer.
Definitely there will be a huge load especially if the number of clients increase.
So there will be a lot of distributed transactions.
I have no doubt whether i should use ejbs or not.
I must because the business logic layer must support remote access.
Considering this scenario, should i use ejbs beyond the session facade layer?
I understand that by using ejbs only in the session facade layer is an adavantage as that will not tie my business logic to a framework.
But considering the scope of the application, kindly provide your suggestions.
Thanks in advance,
James.

Similar Messages

Business Delegate & Session Facade

1 What is the need of Business delegate? For controlling the interactions between the presentation tier and business tier, session facade pattern can be used.
2 Can a session facade have all the facade related methods. (i.e. entire application has only one session facade). In the pattern documentation, it is stated as this stratergy inefficient. Why is so?

1 What is the need of Business delegate? The delegate provides a client-side business abstraction and acts as a proxy for the business service. Note that it is a proxy, not a facade. In other words, it is a "stand-in" for the biz service, but it is not aggregating several method invocations into one and controlling the processing flow. So, the delegate is not providing a higher-level interface to some underlying system, as would a facade.
So what's the need, if it's just a proxy? Obviously, it adds a layer, so one must consider its benefits versus the drawback of writing, maintaining, and interacting with this extra layer. In a sophisticated system, the benefits are many and will usually outweigh the drawbacks. Following are some of the benefits of using a delegate in the web container as an access point for invocations to the business tier:
- It hides the underlying implementation details of the business service (ie: lookup, creation, access details of EJB...The delegate may use the Service Locator pattern to achieve this goal).
- The delegate translates exceptions as necessary. Service-level exceptions, such as java.rmi.Remote can be intercepted by the delegate and translated into application exceptions. This also helps shield web container clients, for example, from EJB and network-related implementation details, which they should not need to worry about.
- The delegate can attempt retries/recovery efforts transparent to its clients, such that this code can be localized and reused, instead of being duplicated in numerous clients.
- The delegate can also manage a cache of resources/results, limiting unecessary remote roundtrips
- The delegate can be used as a control point for multi-threaded access into the business tier. For example, stateful beans are not to be invoked by multiple client threads simultaneously. The delegate can serialize access, or, as mentioned above, handle exceptions/retries, as necessary.
For
controlling the interactions between the presentation
tier and business tier, session facade pattern can be
used.If you decide to use a delegate in your system, based on considering its benefits, as described above, then you'll typically create a one-to-one mapping between a delegate and a session facade. The delegate is not controlling the processing flow per se, but rather serving as a single access point and proxy as described above. The facade acts as a controller, though, aggregating the interfaces of underlying components and subsystems into a higher-level interface. If you find that your using the delegate more as a facade, we recommend pushing this control code from the web container (where the delegate lives) back to the session bean in the ejb container, so that the logic can be reused across clients.
>
2 Can a session facade have all the facade related
methods. (i.e. entire application has only one session
facade). In the pattern documentation, it is stated
as this stratergy inefficient. Why is so?Having a single facade for the entire system would prove to be quite a maintenance and manageability headache for a non-trivial application. Imagine that every time something changes in any subsystem of the service layer, developers are all making changes to the same file and the same class...Not desirable and not very elegant. At the same time, one does not want to go the other way and create too many facades by having a one-to-one mapping of session facades to use cases, for example. We recommend aggregating a group of related interactions into a session facade, so that it provides highly course-grained access and control to underlying components.
Thanks,
Dan

DAO and session Facade

Hi
I came across an design pattern where they call the DAO from Session facade. The purpose of session facade is to minimise the number of network calls. Even though there are no network calls involved why do they go for session facade to call the DAO. Wont a DAOFActory will help them doin this.
Even if the requirement is to block the presentation layer of all the backend operations, we could have had any other , why session facade was used i could nt understand

You might even discover that entity beans can get created even before you use them.
Your application server creates "pools" of bean instances that it can use when it needs to. It is part of this role, and is done in order to optimize performances.
You cannot force them to be garbage collected. Even if you stop referencing them, the app server will.
When your create references to a bean, it (usually) won't create an instance. It will take an existing one, and load data into it, using ejbCreate or ejbActivate.
Hope this helps.
/Stephane

Bussiness Delegate vs Session Facade.

Hi,
Please any one tell me what are the differences between Bussiness Delegate and Session Facade.
I read some documentation in online but not clear in above two design patterns.
Any help I will appricate.
Thanks,
CDulla.

A business delegate is generally thought of as a plug-in that is given to the client developer. This "plug-in" can do service lookups and searches on behalf of the client developer. Very smart if you have a Swing or Servlet Developer that doesn't know ejbs.
A session facade is a session bean that is much like the delegate but resides on the server and is used to encapsulate course services so that network can be held to a minimum.
A business-delegate can call a session facade which in turn can call an entity bean

JDev 1013 BUG? EJB, CMR, DTO and session facade.

I've created two CMP Entity EJB's, CourseEJB and PupilEJB, and defined a M:N relationship between them. A course has many pupils and vice versa. Each bean has only two attributes, id (Long, part of PK) and name (String) for simplicity.
1. Generated DTO's for the two beans won't compile
I get error on:
public void addcourseEJBLocalDTO(courseEJBLocalDTO courseEJBLocalDTO)
coursesDTO.add(courseEJBLocalDTO);
courseEJBLocalDTO.setPupilsDTO(this);
and:
public void addPupilEJBLocalDTO(PupilEJBLocalDTO pupilEJBLocalDTO)
pupilsDTO.add(pupilEJBLocalDTO);
pupilEJBLocalDTO.setCoursesDTO(this);
because JDev has created definitions of the setCoursesDTO function like this:
public void setCoursesDTO(Collection<courseEJBLocalDTO> coursesDTO)
this.coursesDTO = coursesDTO;
and ditto for the setPupilsDTO method, taking a Collection as argument too.
I can fix this code manually, but next step, generating a session Facade with added methods for handling the entity beans isn't exactly any better...
One M:N relation is pretty basic, am I doing something wrong here?
Message was edited by:
jonmarti

Forgot to add, running on 1013

(ejb 3.0) stateful session facade?

Hello,
(1) Would you consider stateful session facade (whose client is a servlet) a bad design?
(2) How to have a stateful session facade referenced in a servlet? I am using Sun Application Server Platform Edition 9.0 and trying to make an injection:
@EJB(name="ejb/KOGSessionFacade")
protected KOGSessionFacade sessionFacade;
but then I get this error in deploy-time: "A stateful session bean can not be injected into a servlet. Please refer to EJB 3.0 Persistence API Specification section #5.2 for further information."
Thanks.

Using a Stateful Session facade from a Servlet is not necessarily a bad design. In some cases, you may want to do some cleanup work on completion of a transaction. Stateful Session beans provide a convenient way (By implementing SessionSynchronization interface) to do this.
In the servlet programming model, Servlets are not designed to be single threaded. So if you inject a StatefulSessionBean (SFSB) references then multiple threads could be possibly be invoking the SFSB concurrently. In this case, the EJBContainer (as required by the specification), will throw a well defined exception.
What you may want to do is to place this SFSB reference into you Servlet session.

Session Facade Pattern without EJB

All of the session facade pattern examples that I can find are heavily emphasized on EJB. I am planning on implementing a system without EJB's, and am interested in the advantages of using the Session Facade pattern.
Are there any good examples of using the Facade pattern outside of EJB to build a java server system? If there aren't, is there a better way to implement this type of strategy in a non-EJB system?

What is the best way to allow the client to
interact
with Entity objects, without exposing them. ShouldI
implement Value objects that are easily serialized
for transfer across the network?
If you are not using EJB or a distributed
architecture, you do not need DTO's or VO's. They
are an odious construct used to obviate the
shortcomings of remote method calls over anetwork.
I strongly disagree.
They are an obvious and easy model used when the
following characteristics of a system exist.
- Different layers.
- Related data items that are moved throughout the
layers.
- Different usage of the related data items by
different layers.
Yes, I agree. Note the qualification 'if you are not using a distributed architecture'.
None of that has anything to do with remote method
calls over a network. Point of fact the first time I
used them was in a stand alone application with no
network traffic except that needed by the database
driver itself.
Simply create your domain model objects as POJO's
s (plain ole Java objects). Persist and retrieve
your objects via DAO's (data access objects).Your
service layer (facade) will sit on top of thedomain
model, encapsulating method calls into logicalunits
of work.
And how exactly do you move the customer
name/customer address from the DAO to the GUI level?I am assuming the dedicated remoting product the OP specifid, Hessian, will handle moving between the tiers, or at least between the controller/view and the business tier.
- Saish

How do I keep a page as my home page when dragging icon to left of URL to house image & confirming that I want it as home page works only for current session, so when Firefox next opened, I end up with some stupid search page which McAfee doesn't like?

A couple of days ago, I connected to the Internet as usual & opened Firefox, only to be greeted by an almost blank screen with a Google-type search box in the middle & this message from McAfee:
"Your default search settings have changed. This may pose a security risk. Would you like to restore them to McAfee Secure Search to provide a safer searching experience?"
instead of my usual home page (BT Yahoo). I naturally clicked on the "Yes" option in the message, but rather than restoring my BT Yahoo home page, all that did was to insert the McAfee logo to the left of the search box in the top right-hand corner of the screen. The first time it happened, I had to search for the BT Yahoo page & then followed the standard procedure for setting it again as my home page. It worked only for that session: each time I shut down Firefox or restarted my computer after that, all I got was the blank "search page" & restoring the previous session was the only way to get back to BY Yahoo. How on earth do I make the home page setting permanent?
As far as I'm aware, I have done nothing to alter my search settings. However, I am anything but computer-literate, so I may have done/pressed something without realising it, but trying to understand what is now happening is far beyond my limited IT skills.

See McAfee support to find out how to disable that McAfee feature - that isn't part of the normal Firefox installation.

I set cookie permissions for a domain to "Allow for session" and at a later time it has changed to "Allow first party only" and now allows persistent cookies.

As it says above, I have cookies set to always ask. I go to a domain (i.e. google) and it asks and I allow for session only. Some time later it will change to "first party only" and start to allow persistent cookies for the domain.

Hello,
Many site issues can be caused by corrupt cookies or cache. In order to try to fix these problems, the first step is to clear both cookies and the cache.
Note: ''This will temporarily log you out of all sites you're logged in to.''
To clear cache and cookies do the following:
#Click the menu button [[Image:New Fx Menu]], choose History, and then Clear Recent History....
#Under "Time range to clear", select "Everything".
#Now, click the arrow next to Details to toggle the Details list active.
#From the details list, check ''Cache'' and ''Cookies'' and uncheck everything else.
#Now click the ''Clear now'' button.
Further information can be found in the [[Clear your cache, history and other personal information in Firefox]] article.
Did this fix your problems? Please report back to us!
Thank you.

How to create a query in your session facade.

Please help me
My code is as follows
public List<TrpSprojects> getProjectNotAssignedToUser(TrpSemployees employee) {
    List proassignements = this.findProjectAssingmentForUser(employee);
    System.out.println("Size of project assignment list :"+proassignements.size());
    List projectIdAssigned = new ArrayList();
    String queryText = "Select o from Trp_Sprojects o where o.project_id not in (";
    for(Iterator proassign = proassignements.iterator();proassign.hasNext();)
        queryText = queryText+((TrpSprojectAssignment) proassign.next()).getTrpSprojects().getProjectId().toString()+",";
    System.out.println(queryText.substring(0,queryText.length()-1)+");");
   String querytoBeExecuted = queryText.substring(0,queryText.length()-1)+")";
    System.out.println("The query to be executed: "+ queryText);
     Query q = em.createNativeQuery(querytoBeExecuted, TrpSprojects.class);
     return q.getResultList();
    }In my session facade i have to create a query that will return the projects that have not been assigned to the user.
When i display the queryText it is as follows:
Select o from Trp_Sprojects o where o.project_id not in (2,1,4,5,12,14,15,16,17,18,20,6,7,8,9,10,11,13,22,19,3,26,21,23)when am executing the query i am getting exception ...
Can anyone help me please . I have not defined this query in the TrpSprojects Class.
Part of my exception its too long am only putting a part
Feb 11, 2008 12:37:07 PM oracle.j2ee.rmi.RMIMessages EXCEPTION_ORIGINATES_FROM_THE_REMOTE_SERVER
WARNING: Exception returned by remote server: {0}
javax.ejb.EJBException: Exception [TOPLINK-4002] (Oracle TopLink Essentials - 2006.8 (Build 060829)): oracle.toplink.essentials.exceptions.DatabaseException
Internal Exception: java.sql.SQLException: ORA-00904: "O": invalid identifier
Error Code: 904
Call:Select o from Trp_Sprojects o where o.project_id not in (2,1,4,5,12,14,15,16,17,18,20,6,7,8,9,10,11,13,22,19,3,26,21,23)
Query:ReadAllQuery(org.trp.model.TrpSprojects); nested exception is:
     Exception [TOPLINK-4002] (Oracle TopLink Essentials - 2006.8 (Build 060829)): oracle.toplink.essentials.exceptions.DatabaseException
Internal Exception: java.sql.SQLException: ORA-00904: "O": invalid identifier
Error Code: 904
Call:Select o from Trp_Sprojects o where o.project_id not in (2,1,4,5,12,14,15,16,17,18,20,6,7,8,9,10,11,13,22,19,3,26,21,23)
Query:ReadAllQuery(org.trp.model.TrpSprojects); nested exception is: oracle.oc4j.rmi.OracleRemoteException: Exception [TOPLINK-4002] (Oracle TopLink Essentials - 2006.8 (Build 060829)): oracle.toplink.essentials.exceptions.DatabaseException
Internal Exception: java.sql.SQLException: ORA-00904: "O": invalid identifier
Error Code: 904
Call:Select o from Trp_Sprojects o where o.project_id not in (2,1,4,5,12,14,15,16,17,18,20,6,7,8,9,10,11,13,22,19,3,26,21,23)
Query:ReadAllQuery(org.trp.model.TrpSprojects); nested exception is:
     Exception [TOPLINK-4002] (Oracle TopLink Essentials - 2006.8 (Build 060829)): oracle.toplink.essentials.exceptions.DatabaseException
Internal Exception: java.sql.SQLException: ORA-00904: "O": invalid identifier
Error Code: 904
oracle.oc4j.rmi.OracleRemoteException: Exception [TOPLINK-4002] (Oracle TopLink Essentials - 2006.8 (Build 060829)): oracle.toplink.essentials.exceptions.DatabaseException
Internal Exception: java.sql.SQLException: ORA-00904: "O": invalid identifier
Error Code: 904
Call:Select o from Trp_Sprojects o where o.project_id not in (2,1,4,5,12,14,15,16,17,18,20,6,7,8,9,10,11,13,22,19,3,26,21,23)
Query:ReadAllQuery(org.trp.model.TrpSprojects)
     at com.evermind.server.ejb.EJBUtils.getUserException(EJBUtils.java:346)
     at com.evermind.server.ejb.interceptor.system.AbstractTxInterceptor.convertAndHandleMethodException(AbstractTxInterceptor.java:69)
     at com.evermind.server.ejb.interceptor.system.TxRequiredInterceptor.invoke(TxRequiredInterceptor.java:52)
     at com.evermind.server.ejb.interceptor.InvocationContextImpl.proceed(InvocationContextImpl.java:119)
     at com.evermind.server.ejb.interceptor.system.DMSInterceptor.invoke(DMSInterceptor.java:52)
     at com.evermind.server.ejb.interceptor.InvocationContextImpl.proceed(InvocationContextImpl.java:119)
     at com.evermind.server.ejb.InvocationContextPool.invoke(InvocationContextPool.java:55)
     at com.evermind.server.ejb.StatelessSessionEJBObject.OC4J_invokeMethod(StatelessSessionEJBObject.java:87)
     at SCTimeReport_RemoteProxy_4p6ko7e.getProjectIdAssignedToUser(Unknown Source)
     at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
     at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
     at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
     at java.lang.reflect.Method.invoke(Method.java:585)
     at com.evermind.server.rmi.RmiMethodCall.run(RmiMethodCall.java:53)
     at com.evermind.util.ReleasableResourcePooledExecutor$MyWorker.run(ReleasableResourcePooledExecutor.java:303)
     at java.lang.Thread.run(Thread.java:595)Regards
Lutchumaya

Thanks for your help Frank.
But when i display my query to be executed by the entityManager it is like that
Select o from Trp_Sprojects o where o.project_id not in (2,1,4,5,12,14,15,16,17,18,20,6,7,8,9,10,11,13,22,19,3,26,21,23)
Regards
Lutchumaya

Business Delegate and Session Facade Pattern

Hi!
The only way to use the Session Facade Pattern, is if I use EJB for Persistence.
Is valid to do this?:
I use Ejb for simple update, insert querys. From my business delegate I call the Session Facade Layer, and from this I invoque Entyties for persistence.
But if I have complex querys, is correct to make PL SQL Procedures and do the following:
From my business delegate I call to the Dao layer, and from this via JDBC I call the Procedure.
Please explain me the best form to do this, having complex querys for reporting and simple querys for inserts/update.
Is valid to combine the use of CMP (for simple persistence cases), BMP (for complex persistence cases), and JDBC for complex select querys with multiple result rows.
Thanks!!

It depends on your design goals. One of the forces driving the use of patterns is the desire to tier an application and abstract the internals of each tier away from the other tiers. One (normal) benefit of this methodology is that the application should become more portable. Now, if you are using PL/SQL, BMP and CMP, you are mixing and matching portable versus proprietary. This is okay. But you should abstract away whether you are using PL/SQL, BMP or CMP from the business layer (or domain model). Ideally, you could completely swap database technologies (say from relational to object) and only have to re-write your integration tier (your DAO's).
Session facade is simply a glorified Facade pattern. It abstracts the lower-level details of what you are persisting and instead forces you to think in terms of objects. You can use a facade without even remotely touching EJB's. It just so happens that, historically, EJB containers made a mess out of scalability when it came to entity beans. The "session" facade refers to a facade pattern implemented in session beans (that have fewer scalability problems) communicating with the entity beans which do have scalability problems.
So... to make a long story short, use whatever persistence mechanism you desire. Remember that you should be able to completely switch your persistent store from one type or another and simply re-write your DAO's (or CMP deployment descriptors). Using a facade pattern can help towards this end, but don't make it a straightjacket.
- Saish
"My karma ran over your dogma." - Anon

Session Facade and Business Delegate Pattern

Hi!
The only way to use the Session Facade Pattern, is if I use EJB for Persistence.
Is valid to do this?:
I use Ejb for simple update, insert querys. From my business delegate I call the Session Facade Layer, and from this I invoque Entyties for persistence.
But if I have complex querys, is correct to make PL SQL Procedures and do the following:
From my business delegate I call to the Dao layer, and from this via JDBC I call the Procedure.
Please explain me the best form to do this, having complex querys for reporting and simple querys for inserts/update.
Is valid to combine the use of CMP (for simple persistence cases), BMP (for complex persistence cases), and JDBC for complex select querys with multiple result rows.
Thanks!!

The session facade is borrowed from the facade pattern, so you could have a facade to almost anything.
A session facade is usually against other beans...for example if you
(Not deployed as CMR)
1. Address bean
2. Phone Bean
3. Customer bean
as entity beans then you can build a session facade such as
CustomerFacadeBean which will CreateReadUpdateDelete(CRUD) address and phone as well when a customer is CRUD ed.
In your case you said you have very complex sql's so one way would be
Business Delegate -> Session Facade -> (Bean Managed) Entity Bean -> DataAccessObject.
If you think this is a overkill then, write a custom pattern such as DataDelegate
DataDelegate -> (Bean Managed) Entity Bean -> DataAccessObject.
In the above scenario you should be quite sure that the system will not evolve into a case mentioned above such as when customer is deleted a phone and address object are also thrown away.
You said this scenario
Business Delegate -> DAO..
I don't think this is a good idea because, a business delegate is a proxy against business objects....and usualy in J2EE business objects are session beans. Entity and DAO objects are data objects not business ones.
Besides if you do the above scenario you will end up wiring all the transactions, etc etc in your DAO which is not a good idea.
Although you could use BusinesDelegates against entity beans it will lead to a misnomer.

Session facade vs Command object pattern

Hello,
I am debating using the Command pattern as my primary strategy for implementing my J2EE enterprise app business logic and would like some advice.
The general idea is to have only a few types of abstract commands (such as a ReadComand, an UpdateCommand -- I might have a slightly finer granularity than this, I don't know) and implement each use case as a command instance. For example, I might have a command called GetOrderItemsCommand which is a ReadCommand and returns the list of order items (i.e., has a getItems() method) when execute()'d.
The result of my design will be a very small set of stateless session beans with an execute() interface. No entity beans (will use Hibernate) and may very well eventually implement a Message bean interface with a similar executeAsynchronously() type method, if eventually necessary.
I guess the popular alternative (or, more correctly, the norm) is to implement session facades for this.
However, I am attracted to the Command pattern because of its simplicity, it's rapid application development angle, its disconnected-ness from EJB (e.g., a modification to a command does not mean touching any EJB at all, just the command class), and the ability for implementing cross-cutting features (such as transactions or security, although I also plan to use an AOP solution at the POJO level). At the same time, it works as a delegate of an EJB, so it's accessible for all kinds of network clients.
I am also familiar with it's drawbacks (e.g., maintainability for large numbers of commands, per the literature, such as the Marinescu 2002 EJB Patterns book).
So, very few types of stateless session beans (though there may be many instances) -- mainly there for remote client acceessibility and the ability to leverage application-level transactions if needed.
Nevertheless, I have never tried the command pattern approach and was curious if others had feedback or case studies.
Best regards --

You normally use Command to decouple your controller from the view and model tiers. Your take on making command objects 'polymorphic' is interesting, and an angle I had not thought of. In general, I implement command objects rather like the Action class in Struts.
Each command has implements an execute() method declared in the ICommad interface. The constructor of the command object ensures that a given command has all the variables and data required to execute properly. I also create a likes() method that returns a boolean. That way, I can add all my commands to a handler and iterate until on returns true on likes().
- Saish

Session Facade design and TopLink objects

Hi, we are just about to start using Toplink as our ORM tool. Ideally we want to wrap all business logic behind stateless session bean facades. So instead of BMP entity beans we are thinking of encapsulating the relevant Toplink objects in each facade and querying, updating that object when the session facade bean is instantiated.
We are using Weblogic 70. Example code:
e.g. MySessionBean implements sessionbean...
     MyTopLinkObject myObject = null;
     public void populateObject(String myID) throws java.rmi.RemoteException {
          try {
               MyTopLinkObject myObject = new MyTopLinkObject(myID);
          catch (Exception e) {
     public void setMyTopLinkObjectsName(String name) {
     //persist this change via the toplink object...
     myObject.setName(name);
Then in MyTopLinkObject:
public class MyTopLinkObject {
     private String myID;
     private String myName;
public MyTopLinkObject(String myID) {
          build(myID);
public void build(String myID) {
          //read TOPLINK cache and populate this MyTopLinkObject name etc
public void setName(String myName) {
     //persist this change...
     this.myName = myName;
// more gets sets
I will then want to update MyTopLinkObject data in for example in calling. MySessionBean.setMyTopLinkObjectsName("my name") from the remote interface on the client.
So what I want to know is how do I can I do this efficiently. I'm too too versant with TopLink caches etc. For example to read the cache do I have to pass thru a TopLink client/server Session to each Session Facade bean so the cache is available for that read, update etc?
Any other tips is using this design would be appreciated!
Thanks in advance,
J

Don, thanks for ur advice yet again.
I think now I am getting a better grasp of how TopLink should be used. However there is very little doco out there in terms of using TOPLInk in as EJB session facade scenario.
After your advice my thoughts are to provide something like the TOPLInkFactory object as you say which will encapsulate the ServerSession TOPLink API and provide more fine grain access to the TOPLInk API. e.g. instead of having to create a ClientSession in each client call b4 getting a UnitOfWOrk, this factory class would provide a method to do this directly:
public final class TOPLinkFactory {
     private static ServerSession serverSession = null;
     private static final String TOPLINK_CONFIG_FILE = "C:\\test.xml";
     private static final String TOPLINK_SESSION = "toplink";
     public TOPLinkFactory() throws Exception {
          super();
     private static ServerSession createServerSession(ClassLoader cl) throws Exception {
          try {
          Project builderProject = XMLProjectReader.read(TOPLINK_CONFIG_FILE, cl);
     serverSession = (ServerSession)builderProject.createServerSession();
          serverSession.login();
               return serverSession;
          catch (Exception e) {
          return null;
     public static UnitOfWork getUnitOfWork(ClassLoader cl) {
          try {
               return createClientSession(cl).acquireUnitOfWork();
     public static ClientSession createClientSession(ClassLoader cl) throws Exception {
          try {
               if (serverSession == null) {
serverSession = createServerSession(cl);
ServerSession s = (ServerSession)SessionManager.getManager().getSession(TOPLINK_SESSION, cl);
               return serverSession.acquireClientSession();
          catch (Exception e) {
          return null;
Then in my Session EJBs I could write a method like this where Thing is a TOPLInk object mapped to the thing table:
     public boolean setThingStatus(String thingID) {
ExpressionBuilder builder = new ExpressionBuilder();
Expression expression = builder.get("THING_ID").equal( Integer.parseInt(thingID));
ClientSession s = TOPLinkFactory.createClientSession(this.getClass().getClassLoader());
Thing thing = (Thing) s.readObject(Thing.class, expression);
UnitOfWork uow = TOPLinkFactory.getUnitOfWork(this.getClass().getClassLoader());
Thing thingClone = (Thing)uow.registerObject(thing);
thingClone.setStatus("WITH JASON");
TOPLinkFactory.commitAndReleaseUnitOfWork(uow);
return true;
Now if the above looks OK and there are no Thread/Session issues that you could make me aware of (or perhaps you would encourage using a full Singleton pattern instead).
I have another question in relation to ClassLoaders. In this architecture I have been passing around the ClassLoader (as above) to the TOPLInkFactory methods to create the unit of works etc... Is this necessary? Or could I avoid it?
Also I think some people in my team are thinking its a good idea for all TOPLink business objects to extend an abstract base class that contains this behaviour and provides these TOPLinkFactory methods. I no expert on TOPLInk but I would have thought this code should not be coupled with the TOPLink business objects for reasons such as:
1) Why should a TOPLink Person object (attributes name, age etc) e.g. extend a class with TOPLink API methods and attributes - I just don't see the logic there as I thought TOPLink was to be non-intrusive. For example a Person object should not encapsulate a ServerSession object and nor does it need to. To this end they are then coupled and if for example we wanted to move these Person objects to work with another persistence framework then they would have to be rewritten to remove all TOPLInk attribute/method references
2) I would have thought accessing the TOPLInk API is only necessary when the client code needs to invoke calls to the database for reads and updates. i.e. that's where you start getting your sessions, unit of works etc. Like in the setThingStatus Session EJB method (i.e. the client code) above and NOT in your business methods of for example the TOPLink Person object?
Please tell me if I am mistaken!
If you could clear some of this up it would be great.
Regards,
Jason

[HOWTO] JSF + Session Facade + Entity Bean (EJB3)

I'm trying to find a tutorial that shows the advisable way to use JSF with EJB 3 entity beans using the Session Facade pattern (?), but the only tutorials around show very basic usage of JSF (validation, conversions etc.).
The only thing that is close is this: http://forum.java.sun.com/thread.jspa?threadID=5122284 but still not what I'm looking for.
Is there any documentation available out there..?
Message was edited by:
HakelRausd

Hi,
I updated the online doc in blueprints solutions catalog with a small explanation of using JSF with EJB Session Bean at https://blueprints.dev.java.net/bpcatalog/ee5/persistence/ejbfacade.html
I will add an image soon too. The basic idea for JSF and EJB3 programming model is
JSF Page --> JSF Managed Bean(backing bean) --> Session Bean --> Java Persistence Object.
In general, ther would be several JSF pages,and several JSF Managed Beans, and several Java Persistence Objects, but fewer Session Beans. The Session Bean would be expected to be shared and accessed by several managed beans with with Session Bean serving as a facade to the Java Persistence Objects, so there would be few Session Beans.
The Session Bean would usually be a local Session bean and would only be remote if-and-only-if you wanted RMI-IIOP access. Usually web components like Servlets and JSF Managed Beans use local access to the Session Beans and the access is within the same VM, no remoteness.
Most of the time, the Session Bean would be a Stateless Session Bean, and would not be a Stateful Session Bean.
hth,
Sean

EJBs only for session facade?

Similar Messages

Maybe you are looking for