In-memory-replace for stateful session bean

When I add "<replication-type>InMemory</replication-type>" at the
          weblogic-ejb-jar.xml file,
          the client can't call this stateful session bean and happen exception.
          If I delete "<replication-type>InMemory</replication-type>" at the
          weblogic-ejb-jar.xml, the client can call this ejb.
          How does the In-memory-replac set?
          Is there any limitations about using the In-memory-replace?
          The weblogic-ejb-jar.xml and the exception are list.
          The weblogic-ejb-jar.xml:
          <?xml version="1.0"?>
          <!DOCTYPE weblogic-ejb-jar PUBLIC '-//BEA Systems, Inc.//DTD WebLogic 6.0.0
          EJB//EN' 'http://www.bea.com/servers/wls600/dtd/weblogic-ejb-jar.dtd'>
          <weblogic-ejb-jar>
          <weblogic-enterprise-bean>
          <ejb-name>test_stateful</ejb-name>
          <stateful-session-descriptor>
          <stateful-session-clustering>
          <home-is-clusterable>True</home-is-clusterable>
          <home-load-algorithm>round-robin</home-load-algorithm>
          <home-call-router-class-name>beanRouter</home-call-router-class-name>
          <replication-type>InMemory</replication-type>
          </stateful-session-clustering>
          </stateful-session-descriptor>
          <jndi-name>test_stateful</jndi-name>
          </weblogic-enterprise-bean>
          </weblogic-ejb-jar>
          The exception:
          java.rmi.RemoteException: EJB Exception: ; nested exception is:
          java.lang.ClassCastException:
          weblogic.rmi.internal.BasicRequestDispatcher
          java.lang.ClassCastException: weblogic.rmi.internal.BasicRequestDispatcher
          <<no stack trace available>>

This is a Work as designed in my opinion
with NRU ,the container work with an eager remove algorithm, this as stated need pressure at the cache to push the container to remove the beans from the cache ,hence there will no passivation unless the max-beans-in-cache is reached, hence you can try reducing the max-beans-in-cache and see how the caching behaves , if It does not help ,chnage the Algorithm of caching or open a Service Request for us , we can look at this closely
Regards
Anis

Similar Messages

What's the difference between the LRU and NRU strategy for stateful session bean?

Hi,
Does anybody know the difference between the LRU and NRU strategy for
stateful session bean?
Thanks,
Levi

LRU makes the assumption that the bean that has been used a lot recently is
likely to be used again.
NRU simply removes the beans that have not been used for a stipulated amount
of time.
The algorithm for the LRU is much more complicated than the NRU. I think BEA
recommends LRU for better performance and NRU when you have memory
constraints.
"levi" <[email protected]> wrote in message
news:[email protected]..
Hi,
Does anybody know the difference between the LRU and NRU strategy for
stateful session bean?
Thanks,
Levi

Max number of instances for stateful session beans

Hi,
I am using OC4J 9.0.3 which is embedded in JDeveloper. I would like to set the maximum number of instances for stateful session beans. I know that the session-deployment section of the orion-ejb-jar.xml file has an attribute called max-instances, but it only works for stateless session beans. (at least it says so in the documentation and it did not work with my stateful session bean)
Anyone knows how to do it?
Thanks,
Leonardo Penha

Leonardo -- Stateful session beans do not come from a pool but are assigned 1 bean to 1 user. When the EJBs life ends the bean goes away. If what you are asking about is passivation configuraiton then that is different. We currently do not support passivation based on some count of SFSBs that exists inside the container but we will provide that capability in the next release of OC4J. (Note: When passivation occurs is mostly up to the vendors and it is not required by the J2EE 1.3 specification based on performance metrics. What is required is that when a SFSB is passivated it is done correctly and we do that.)
Thanks -- Jeff

Accessing the same stateful session bean from multiple clients in a clustered environment

I am trying to access the same stateful session bean from multiple
          clients. I also want this bean to have failover support so we want to
          deploy it in a cluster. The following description is how we have tried
          to solve this problem, but it does not seem to be working. Any
          insight would be greatly appreciated!
          I have set up a cluster of three servers. I deployed a stateful
          session bean with in memory replication across the cluster. A client
          obtains a reference to an instance of one of these beans to handle a
          request. Subsequent requests will have to use the same bean and could
          come from various clients. So after using the bean the first client
          stores the handle to the bean (actually the replica aware stub) to be
          used by other clients to be able to obtain the bean. When another
          client retrieves the handle gets the replica aware stub and makes a
          call to the bean the request seems to unpredictably go to any of the
          three servers rather than the primary server hosting that bean. If the
          call goes to the primary server everything seems to work fine the
          session data is available and it gets backed up on the secondary
          server. If it happens to go to the secondary server a bean that has
          the correct session data services the request but gives the error
          <Failed to update the secondary copy of a stateful session bean from
          home:ejb20-statefulSession-TraderHome>. Then any subsequent requests
          to the primary server will not reflect changes made on the secondary
          and vice versa. If the request happens to go to the third server that
          is not hosting an instance of that bean then the client receives an
          error that the bean was not available. From my understanding I thought
          the replica aware stub would know which server is the primary host for
          that bean and send the request there.
          Thanks in advance,
          Justin


          If 'allow-concurrent-call' does exactly what you need, then you don't have a problem,
          do you?
          Except of course if you switch ejb containers. Oh well.
          Mike
          "FBenvadi" <[email protected]> wrote:
          >I've got the same problem.
          >I understand from you that concurrent access to a stateful session bean
          >is
          >not allowed but there is a
          >token is weblogic-ejb-jar.xml that is called 'allow-concurrent-call'
          >that
          >does exactly what I need.
          >What you mean 'you'll get a surprise when you go to production' ?
          >I need to understand becouse I can still change the design.
          >Thanks Francesco
          >[email protected]
          >
          >"Mike Reiche" <[email protected]> wrote in message
          >news:[email protected]...
          >>
          >> Get the fix immediately from BEA and test it. It would be a shame to
          >wait
          >until
          >> December only to get a fix - that doesn't work.
          >>
          >> As for stateful session bean use - just remember that concurrent access
          >to
          >a stateful
          >> session bean is not allowed. Things will work fine until you go to
          >production
          >> and encounter some real load - then you will get a surprise.
          >>
          >> Mike
          >>
          >> [email protected] (Justin Meyer) wrote:
          >> >I just heard back from WebLogic Tech Support and they have confirmed
          >> >that this is a bug. Here is their reply:
          >> >
          >> >There is some problem in failover of stateful session beans when its
          >> >run from a java client.However, it is fixed now.
          >> >
          >> >The fix will be in SP2 which will be out by december.
          >> >
          >> >
          >> >Mike,
          >> >Thanks for your reply. I do infact believe we are correctly using
          >a
          >> >stateful session bean however it may have been misleading from my
          >> >description of the problem. We are not accessing the bean
          >> >concurrently from 2 different clients. The second client will only
          >> >come into play if the first client fails. In this case we want to
          >be
          >> >able to reacquire the handle to our stateful session bean and call
          >it
          >> >from the secondary client.
          >> >
          >> >
          >> >Justin
          >> >
          >> >"Mike Reiche" <[email protected]> wrote in message
          >news:<[email protected]>...
          >> >> You should be using an entity bean, not a stateful session bean
          >for
          >> >this application.
          >> >>
          >> >> A stateful session bean is intended to be keep state (stateful)
          >for
          >> >the duration
          >> >> of a client's session (session).
          >> >>
          >> >> It is not meant to be shared by different clients - in fact, if
          >you
          >> >attempt to
          >> >> access the same stateful session bean concurrently - it will throw
          >> >an exception.
          >> >>
          >> >> We did your little trick (storing/retrieving handle) with a stateful
          >> >session bean
          >> >> on WLS 5.1 - and it did work properly - not as you describe. Our
          >sfsb's
          >> >were not
          >> >> replicated as yours are.
          >> >>
          >> >> Mike
          >> >>
          >> >> [email protected] (Justin Meyer) wrote:
          >> >> >I am trying to access the same stateful session bean from multiple
          >> >> >clients. I also want this bean to have failover support so we want
          >> >to
          >> >> >deploy it in a cluster. The following description is how we have
          >tried
          >> >> >to solve this problem, but it does not seem to be working. Any
          >> >> >insight would be greatly appreciated!
          >> >> >
          >> >> >I have set up a cluster of three servers. I deployed a stateful
          >> >> >session bean with in memory replication across the cluster. A client
          >> >> >obtains a reference to an instance of one of these beans to handle
          >> >a
          >> >> >request. Subsequent requests will have to use the same bean and
          >could
          >> >> >come from various clients. So after using the bean the first client
          >> >> >stores the handle to the bean (actually the replica aware stub)
          >to
          >> >be
          >> >> >used by other clients to be able to obtain the bean. When another
          >> >> >client retrieves the handle gets the replica aware stub and makes
          >> >a
          >> >> >call to the bean the request seems to unpredictably go to any of
          >the
          >> >> >three servers rather than the primary server hosting that bean.
          >If
          >> >the
          >> >> >call goes to the primary server everything seems to work fine the
          >> >> >session data is available and it gets backed up on the secondary
          >> >> >server. If it happens to go to the secondary server a bean that
          >has
          >> >> >the correct session data services the request but gives the error
          >> >> ><Failed to update the secondary copy of a stateful session bean
          >from
          >> >> >home:ejb20-statefulSession-TraderHome>. Then any subsequent requests
          >> >> >to the primary server will not reflect changes made on the secondary
          >> >> >and vice versa. If the request happens to go to the third server
          >that
          >> >> >is not hosting an instance of that bean then the client receives
          >an
          >> >> >error that the bean was not available. From my understanding I
          >thought
          >> >> >the replica aware stub would know which server is the primary host
          >> >for
          >> >> >that bean and send the request there.
          >> >> >
          >> >> >Thanks in advance,
          >> >> >Justin
          >>
          >
          >

What are the advantages stateful session beans takes over than HttpSession

Hi
We can use HttpSession to maintain the conversational state with the client.
Then why we need to go for stateful session beans to maintain conversational state with the client.
What are advantages we can have while using stateful session beans ratherthan HttpSession????.
Regards
Dhinesh kumar R

I think we can use the magic word in the software development "seperation of concerns" ;-)
HttpSession is in the Servlet/JSP tier and is mainly used in the presentation logic tier (UI, page flows, navigation etc..)
where the stateful session bean's concern is to maintain a conversational state in a business logic tier.
Through this seperation its possible, that the business tier's state is undependant from the HttpSession and you dont create a hardlink that could cause problems later. Imagine having a desktop client accessing your business logic etc. not every client needs to be a web client ;-)
I hope that the answer helps you understanding the topic :-)
Brgds,
Nail

SAME EJB 3.0 Stateful Session bean for different JSP sessions returned

Forum,
I have a strange problem utilizing stateful session beans. Please note, that I am using jdeveloper for the following:
1) Here is a basic stateful session bean:
@Stateful(name = "DemoClass")
public class DemoClassBean implements DemoClass,
           DemoClassLocal, SessionSynchronization {
      public static int id=0;
      public DemoClassBean() {
      id++;
      public int getId() {
         return id;
      //... other methods
}2) This bean is accessed from a JSP for testing purpose, I am copying only the script used in JSP:
<%
   DemoClass bean = null;
   try {
      bean = (DemoClass) ((new javax.naming.InitialContext()).
                   lookup("DemoClass "));
      System.out.println("ID=" + bean.getId() );
      } catch (javax.naming.NamingException e) {
         // TODO
%>If this page is addressed by different clients, from different browsers, the same bean is returned.
Here is what I see in the logs:
ID=1
ID=1
ID=1
The same problem is being observed when this session bean is accessed from a JSF Backing bean.
What could be wrong? Is this a bug in oc4j / jdeveloper (version 10.1.3.3.0)?
Edited by: smw000000001 on Nov 17, 2008 10:52 AM

Hi,
The code for stateful is perfectly fine and working in a normal way. The way you are trying to implement the stateful session bean in your application is wrong.
think of binding the stateful session bean with HttpSession object.
So that you will get a unique stateful session bean object.

Problem using application client for local stateful session bean

Hi,
I have deployed a local stateful session bean in Sun J2EE 1.4 application server.
On running the applclient for the stateful session bean application client i get the following error:
Warning: ACC006: No application client descriptor defined for: [null]
cant we use application client for local stateful session beans. becoz the application runs smoothly when i changed the stateful sesion bean to remote.

Hi,
No, an ejb that exposes a local view can only be accessed by an ejb or web component packaged within the same application. Parameters and return values for invocations through the ejb local view are passed by reference instead of by value. That can't work for an application client since it's running in a separate JVM.
--ken

Stateful Session Beans are not passivated / serialized when cache idle time

Technology: Sun Application Server version 7.0.0_01; JDK 1.4.1; developed on Windows 2000; Tested on Sun Solaris.
Initial error on Sun Solaris:
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr: Exception in thread "service-j2ee-25" org.omg.CORBA.OBJ_ADAPTER: vmcid: SUN minor code: 1015 completed: No
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.POA.GenericPOAServerSC.preinvoke(GenericPOAServerSC.java:389)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.POA.ServantCachePOAClientSC.initServant(ServantCachePOAClientSC.java:112)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.POA.ServantCachePOAClientSC.setOrb(ServantCachePOAClientSC.java:95)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream_1_0.createDelegate(CDRInputStream_1_0.java:760)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream_1_0.internalIORToObject(CDRInputStream_1_0.java:750)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream_1_0.read_Object(CDRInputStream_1_0.java:669)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream_1_0.read_abstract_interface(CDRInputStream_1_0.java:890)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream_1_0.read_abstract_interface(CDRInputStream_1_0.java:884)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream.read_abstract_interface(CDRInputStream.java:307)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.IIOPInputStream.readObjectDelegate(IIOPInputStream.java:228)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.IIOPInputStream.readObjectOverride(IIOPInputStream.java:381)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at java.io.ObjectInputStream.readObject(ObjectInputStream.java:318)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.enterprise.iiop.IIOPHandleDelegate.getStub(IIOPHandleDelegate.java:58)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.enterprise.iiop.IIOPHandleDelegate.readEJBObject(IIOPHandleDelegate.java:38)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.ejb.portable.HandleImpl.readObject(HandleImpl.java:91)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.IIOPInputStream.readObject(Native Method)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.IIOPInputStream.invokeObjectReader(IIOPInputStream.java:1298)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.IIOPInputStream.inputObject(IIOPInputStream.java:908)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.IIOPInputStream.simpleReadObject(IIOPInputStream.java:261)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.ValueHandlerImpl.readValueInternal(ValueHandlerImpl.java:247)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.se.internal.io.ValueHandlerImpl.readValue(ValueHandlerImpl.java:209)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream_1_0.read_value(CDRInputStream_1_0.java:981)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.iiop.CDRInputStream.read_value(CDRInputStream.java:287)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.sun.corba.ee.internal.javax.rmi.CORBA.Util.copyObject(Util.java:598)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at javax.rmi.CORBA.Util.copyObject(Util.java:314)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.telstra.nodeman.ejb._NodeMaint_Stub.getHandle(Unknown Source)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.telstra.nodeman.arch.NMAViewBeanProxy.checkBeans(NMAViewBeanProxy.java:631)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.telstra.nodeman.view.html.NMAStandardButton.handleRequest(NMAStandardButton.java:143)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.telstra.nodeman.arch.NMAViewBeanBase.handleRequest(NMAViewBeanBase.java:1573)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.jato.view.ViewBeanBase.dispatchInvocation(ViewBeanBase.java:824)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.jato.view.ViewBeanBase.invokeRequestHandlerInternal(ViewBeanBase.java:637)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.jato.view.ViewBeanBase.invokeRequestHandler(ViewBeanBase.java:595)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.jato.ApplicationServletBase.dispatchRequest(ApplicationServletBase.java:772)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.jato.ApplicationServletBase.processRequest(ApplicationServletBase.java:446)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.jato.ApplicationServletBase.doPost(ApplicationServletBase.java:324)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.telstra.nodeman.view.ViewServlet.doPost(ViewServlet.java:243)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at javax.servlet.http.HttpServlet.service(HttpServlet.java:760)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at javax.servlet.http.HttpServlet.service(HttpServlet.java:853)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardWrapperValve.invokeServletService(StandardWrapperValve.java:720)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardWrapperValve.access$000(StandardWrapperValve.java:118)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardWrapperValve$1.run(StandardWrapperValve.java:278)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at java.security.AccessController.doPrivileged(Native Method)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardWrapperValve.invoke(StandardWrapperValve.java:274)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:505)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardContextValve.invoke(StandardContextValve.java:212)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:505)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:203)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:505)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.ias.web.connector.nsapi.NSAPIProcessor.process(NSAPIProcessor.java:158)
[10/Aug/2004:08:04:57] WARNING (17227): CORE3283: stderr:      at com.iplanet.ias.web.WebContainer.service(WebContainer.java:598)
The above error caused the server to use all available memory and required a reboot to proceed.
Subsequent testing against the Sun Appliucation Server 7 on Windows 2000 dev environment using the Sun Studio IDE for debugging and trace statements inserted in the code indicate that the Application Server is removing the Stateful Session Beans when they time out without an ejbPassivate event and without serializing the beans to the data-store. cache-idle-timeout-in-seconds set to 180 and removal-timeout-in-seconds set to 1800.
The server.log indicates that the beans are timing out:
[19/Aug/2004:18:15:10] WARNING ( 1664): [NRU-com.telstra.nodeman.ejb.AddressMaintBean]: IdleBeanCleanerTask finished after removing 2 idle beans
Trace statements inserted in ejbPassivate do not appear in the log.
It is my understanding that the above timeout should have caused an ejbPasssivate and serialization of the beans.
The beans have been validated using Sun Java Studio Enterprise 6 with 'EJB validate'.
My reading of the problem is that the beans are not being serialized and the error occurs when the application attempts to reference (getHandle) the bean after timeout.
Any suggestions would be appreciated.

Thanks Thorick.
I am using NRU caching. WL 7.0 SP2.
I have not defined idle-timeout-seconds in my weblogic-ejb-jar.xml. As I understand
the default value for this is 600secs. So the ejbs should be removed after this
time. Below is the
weblogic-ejb-jar.xml that I am using.
<!DOCTYPE weblogic-ejb-jar PUBLIC '-//BEA Systems, Inc.//DTD WebLogic 7.0.0 EJB//EN'
'http://www.bea.com/servers/wls700/dtd/weblogic-ejb-jar.dtd'>

<weblogic-ejb-jar>
<weblogic-enterprise-bean>
<ejb-name>Cart</ejb-name>
<stateful-session-descriptor>
<stateful-session-clustering>
<home-is-clusterable>true</home-is-clusterable>
<replication-type>InMemory</replication-type>
</stateful-session-clustering>
</stateful-session-descriptor>
<transaction-descriptor>
     <trans-timeout-seconds>
          60
     </trans-timeout-seconds>
</transaction-descriptor>
<jndi-name>CartHome</jndi-name>
</weblogic-enterprise-bean>
</weblogic-ejb-jar>
"thorick" <[email protected]> wrote:
>
The idle-timeout-seconds property controls the timeout/removal behavior.
which stateful session cache type are you using ? LRUCache or NRUCache

Help!!! Stateful session bean clustering

          I posted this message on interest.clustering.in-memory-replication. But there is
          no answer at all. Could anyone give me
          some hints? Thanks a lot. I am a student and this work is part of my course project.
          It is due soon.
          =================================================================
          I have a big problem to get stateful session bean in-memory replication run on
          clustering. I have three server instances on three nodes and there is always errors
          from the server instance like this when I run the client to invoke the bean methods:
          <Jan 4, 2002 6:47:31 PM EST> <Error> <EJB> <Failed to update the secondary copy
          of a stateful session bean from home:stateful.ejb.AgentSTF>
          Besides, the load balancing can not work probably too. Below is what I get.
          Statistics for different servers:
          "myserver" processed 160 (12%) of 1311 invocations "ejbperf-3" processed 1151
          (88%) of 1311 invocations
          End Client...
          Load balancing works fine with stateless and entity bean.
          Any comments are greatly appreciated.
          Jenny


Hi Jenny,
          Please could you open a case to the customer support ([email protected]) There is a Cr
          on this issue CR059046
          Description of the CR
          WLS 6.1SP1 - Accessing Stateful Session Bean from Stateless Session Bean in a
          cluster fails with errors
          <Oct 5, 2001 5:08:25 PM GMT+00:00> <Error> <EJB> <Failed to update the
          secondary copy of a stateful session bean from home:UserState>
          Regards
          Stephane
          Jenny Liu wrote:
          > I posted this message on interest.clustering.in-memory-replication. But there is
          > no answer at all. Could anyone give me
          > some hints? Thanks a lot. I am a student and this work is part of my course project.
          > It is due soon.
          >
          > =================================================================
          >
          > I have a big problem to get stateful session bean in-memory replication run on
          > clustering. I have three server instances on three nodes and there is always errors
          > from the server instance like this when I run the client to invoke the bean methods:
          >
          > <Jan 4, 2002 6:47:31 PM EST> <Error> <EJB> <Failed to update the secondary copy
          > of a stateful session bean from home:stateful.ejb.AgentSTF>
          >
          > Besides, the load balancing can not work probably too. Below is what I get.
          >
          > Statistics for different servers:
          > "myserver" processed 160 (12%) of 1311 invocations "ejbperf-3" processed 1151
          > (88%) of 1311 invocations
          > End Client...
          >
          > Load balancing works fine with stateless and entity bean.
          >
          > Any comments are greatly appreciated.
          >
          > Jenny
          >
          Regards,
          Stephane Kergozien
          BEA Support

Local Stateful Session Bean State Failover Cluster

          Hi,
          How are you doing? I have poured through the previous postings and still am unclear
          on the following: If we have a stateful session bean that is local is the state
          replicated in a cluster if we are using in-memory replication?
          What controls when the state is replicated? Transactions?
          What happens if the methods are not transacted? Is the state replication at the
          method level?
          Thank you so much,
          -Bart Simpson


Hi Bart,
          I'm a bit unclear on it too. It's been asked before, but no definite answer
          has been provided.
          Peace,
          Cameron Purdy
          Tangosol, Inc.
          http://www.tangosol.com/coherence.jsp
          Tangosol Coherence: Clustered Replicated Cache for Weblogic
          "Bart Simpson" <[email protected]> wrote in message
          news:[email protected]..
          >
          > Hi,
          >
          > How are you doing? I have poured through the previous postings and still
          am unclear
          > on the following: If we have a stateful session bean that is local is the
          state
          > replicated in a cluster if we are using in-memory replication?
          >
          > What controls when the state is replicated? Transactions?
          >
          > What happens if the methods are not transacted? Is the state replication
          at the
          > method level?
          >
          > Thank you so much,
          > -Bart Simpson

HttpSession vs. Stateful Session Bean ---- when State Session is large

I hope most of the people come across with this issue where to put the state
for the internet/intranet based applications when they are using
servlet/jsps calling session beans. Weblogic 4.5.1 does support httpsession
in-memory replication for the servlets but the stateful session beans are
not replicated in clustered environment. Plus with stateful bean u get
activation/passivation overhead too. So one tempts to use stateless session
beans and store state in httpsession?
Then what is the upper limit for the session state one can put in
HttpSession with the weblogic? Is there any way to configure it?
One way to overcome the httpsession size limitation is to use database for
storing state and just store some unique Ids for the session info in
httpSession. But then there will be overhead for database connection?(jdbc
connection pool can provide some help here). So what is the recommended way
for doing this provided few thousand concurrent clients with session size
say exceeding 4kb per client?
Thanks
Usmani

There are no special setting in sun-ejb-jar.xml regarding cache settings. The default settings from server.xml are used:
    <jdbc-connection-pool steady-pool-size="8" max-pool-size="32" max-wait-time-in-millis="60000" pool-resize-quantity="2" idle-timeout-in-seconds="300" is-isolation-level-guaranteed="false" is-connection-validation-required="false" connection-validation-method="auto-commit" fail-all-connections="false" datasource-classname="oracle.jdbc.pool.OracleDataSource" name="ebs">
      <property value="jdbc:oracle:thin:@myebsdbsserver:1521:ebsdevdb" name="url"/>
      <property value="ebs" name="user"/>
      <property value="ebs" name="password"/>
    </jdbc-connection-pool>
<ejb-container steady-pool-size="32" pool-resize-quantity="16" max-pool-size="64" cache-resize-quantity="32" max-cache-size="512" pool-idle-timeout-in-seconds="600" cache-idle-timeout-in-seconds="600" removal-timeout-in-seconds="5400" victim-selection-policy="nru" commit-option="B" monitoring-enabled="true">
     </ejb-container>The Session Bean uses Container Managed Transaction. Is it possible in this case, that the bean isn't 'idle enough' in order to set into passivated?

Stateful session bean time out

How to specify a stateful session beans's time out value?
Thanks,
Jack

Bear in mind that setting a timeout for a stateful session bean will
typically have no more effect than setting a maximum number of beans. As
with garbage collection in Java, beans become eligible for passivation or
removal when their timeout expires and there are more beans than the
maximum, but they will only be removed or passivated when the server runs
low on memory.
Dave
"jack" <[email protected]> wrote in message
news:[email protected]..
>
How to specify a stateful session beans's time out value?
Thanks,
Jack

Plz help!! idempotent methods stateful session beans? (failover)

          plz help!! idempotent methods stateful session beans? (failover)
          Hi there,
          Im trying to failover my shopping cart implemented using stateful ejb, the replication
          is working but the failover is not (Object not found exception....).
          By design is hard to think on an stateful ejb using idempotent methods, but after
          checking the documentation in detail im completely confused, so im starting to
          wonder if i should use idempotent methods or manual retry in case one server is
          down.
          Plz take a loo at these explanations about faiolver in the WLS edocs.
          "....With clustered objects, automatic failover generally occurs only in cases
          where the object is idempotent.....Because of this, replica-aware stubs will not,
          by default, attempt to retry a method that fails after the request is sent but
          before it returns. This behavior can be overridden by marking a service idempotent..."
          BUT
          "By default, a stateful session bean's Home stub provides load balancing and failover
          for its method invocations to any clustered server where the bean is deployed...."
          So is possible to achieve automatic failover (not only load balancing) for stateful
          ejb without having to worry about state-changes or manual retry. (i know the retry
          is requiered if the server crashed just before finishing a transaction but this
          is not what im talking about).
          Any help will be highly appreciated
          Alan


          Ryan,
          Yes, im deploying using In-memory replication and clusterable elements, when try
          to invoke the method (after one of the nodes is down) i recieve an "java.rmi.NoSuchObjectException:
          Bean has been deleted".
          I believe all this could be related to some sort of issue with JNDI Lookup in
          the cluster, but haavent been able to get a clue to make it work....(almost desesperate
          by now)
          Alan
          "ryan upton" <[email protected]> wrote:
          >Alan,
          >
          >If the state isn't maintained during failover I would suspect a
          >configuration error. Have you set the <replication-type> element within
          ><stateful-session-clustering> tag in the bean's deployment descriptor?
          >
          >"Alan" <[email protected]> wrote in message
          >news:40e17cf7$1@mktnews1...
          >>
          >> Ryan,
          >>
          >>
          >> Maybe i didnt explain myself, in the scenario im dealing with I KNOW
          >the
          >method
          >> finished succesfully (example, the user succesfully added an item to
          >a
          >cart).
          >>
          >>
          >> Just after that one of the nodes goes down, and the user decides to
          >add a
          >new
          >> item (new invocation , not a retry of failed one) on the same session.
          >>
          >> This is the case in which failover should work automatically as you
          >say,
          >meaning
          >> the new method invokation should be able to retrieve the session state
          >info from
          >> the replica and switch the primary.....Well, this is the part wich
          >isnt
          >working
          >> :( as i understand it should transparently
          >>
          >> I believe could be a config issue , but have no way to make it work.
          >>
          >> Regards,
          >>
          >> Alan
          >> "ryan upton" <[email protected]> wrote:
          >> >
          >> >"Alan" <[email protected]> wrote in message
          >news:40e06296@mktnews1...
          >> >>
          >> >> Ryan,
          >> >>
          >> >> Thanks for clearing the both conditions a bit more, in this case
          >my
          >> >issue
          >> >has
          >> >> to do with the scenario #1 in which subsequent calls (invocations)
          >> >go to a
          >> >node
          >> >> which isnt avaliable(down), do you have any idea of in what conditions
          >> >the
          >> >"automnatic"
          >> >> failover would fail??? (meaning why the stub would try to dispatch
          >> >a call
          >> >to the
          >> >> same previous node)
          >> >>
          >> >> Thanks for the help again
          >> >>
          >> >
          >> >Subsequent calls won't go to a node that's down. That's the whole
          >> >point
          >> >;-). Failover is always automatic, what you are getting confused
          >on
          >> >is
          >> >automatic method re-execution upon failover which happens if the method
          >> >has
          >> >been marked as idempotent. The logic is this: if I failover to an
          >EJB
          >> >that
          >> >I know is still alive, I can't safely call the same method again.
          > Why?
          >> >Because I don't know how much of the method completed the last time
          >before
          >> >the previous bean's failure. So if the bean can yield multiple results
          >> >from
          >> >multiple invocations I had better play it safe and not re-execute
          >the
          >> >method
          >> >on the bean I failed over to (unless it's idempotent because in that
          >> >case it
          >> >doesn't matter how many times I call it, it always does the same thing).
          >> > In
          >> >scenario #1 and #2 failover is automatic and all new calls are executed
          >> >against the bean the client failed over to, but only idempotent methods
          >> >are
          >> >re-executed.
          >> >
          >> >
          >>
          >
          >

How create EJB 2.1 Stateful Session Bean in a EJB 3.0 Session Bean

Hi All,
We have been developing on EJB 2.1. We are now adding a module on EJB 3.0.
How can we "create" a stateful session bean with create method signature similar to create(String id)?
We have tried
// this is the remote interface
@EJB AddressBean abean;
But not working
Any help will be appreciated.

There is no explicit create() call for EJB 3.0 session beans. It doesn't really matter though
since you can accomplish the same thing by defining your own business method to act
as an initializer for whatever state you'd like. E.g.
@Remote
public interface FooInterface {
public void initialize(String id);
// ... other business methods
@EJB FooInterface fooRef;
fooRef.initialize("foo");
--ken

EJB 3.0 Stateful session bean shared between Servlet's

Hello
I have a bit of a noob question regarding Stateful sessions beans.
I am wanting to know if there is a way that I can share an instance of a session bean between multiple HttpServlet instances?
I am sending XML messages from a mobile J2ME application, there will be several http POST's made from the mobile client to the server. I would like these multiple POST's to be passed from the handling servlet instance to the same uniquely identified single stateful session bean instance (i can then @Remove the stateful bean when I have finished my several requests).
I would greatly appreciate any tips anyone could give me.

If not, your only option is to maintain the
association yourself by creating a unique id for
each
conversation and storing that id along with the SFSB
reference in the ServletContext. Then you'll
need to pass the id in along with each invocation to
retrieve the appropriate SFSB reference.Thanks for your reply.
Will I always be presented with the same ServletContext instance? Even if the time between requests might be many minutes? Where can I learn more about how to use the ServletContext?
Thanks!

In-memory-replace for stateful session bean

Similar Messages

Maybe you are looking for