https://issues.jboss.org This file is an XML representation of an issue en-us 6.3.15 6346 03-02-2015 https://issues.jboss.org/browse/SEAMJMS-16 Seam JMS <p>While working on seam jms, I've noted the following behavior.</p> <p>java.lang.IllegalStateException: Singleton not set for BaseClassLoader@3479404a</p> {vfs:///apps/jboss/jboss-6.0.0.Final/server/all/conf/jboss-service.xml} <p> at org.jboss.weld.integration.provider.JBossSingletonProvider$EarSingleton.get(JBossSingletonProvider.java:59) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.Container.instance(Container.java:58) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.resolution.ResolvableBuilder.checkQualifier(ResolvableBuilder.java:209) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.resolution.ResolvableBuilder.addQualifier(ResolvableBuilder.java:174) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.resolution.ResolvableBuilder.addQualifierIfAbsent(ResolvableBuilder.java:184) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.manager.BeanManagerImpl.resolveObserverMethods(BeanManagerImpl.java:464) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.manager.BeanManagerImpl.fireEvent(BeanManagerImpl.java:604) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.weld.manager.BeanManagerImpl.fireEvent(BeanManagerImpl.java:598) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.jboss.seam.jms.bridge.IngressMessageListener.onMessage(IngressMessageListener.java:45)<br/> at org.hornetq.jms.client.JMSMessageListenerWrapper.onMessage(JMSMessageListenerWrapper.java:91) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.hornetq.core.client.impl.ClientConsumerImpl.callOnMessage(ClientConsumerImpl.java:822) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.hornetq.core.client.impl.ClientConsumerImpl.access$100(ClientConsumerImpl.java:46) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.hornetq.core.client.impl.ClientConsumerImpl$Runner.run(ClientConsumerImpl.java:940) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at org.hornetq.utils.OrderedExecutorFactory$OrderedExecutor$1.run(OrderedExecutorFactory.java:100) <span class="error">&#91;:6.0.0.Final&#93;</span><br/> at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) <span class="error">&#91;:1.6.0_22&#93;</span><br/> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) <span class="error">&#91;:1.6.0_22&#93;</span><br/> at java.lang.Thread.run(Thread.java:662) <span class="error">&#91;:1.6.0_22&#93;</span></p> <p>This exception occurs when trying to fire an event with the contents of a JMS message. The message listener in this case is instantiated from the portable extension and receives the bean manager as an argument in the constructor. I am able to get as far as reading the data from JMS, but when firing the second event the above exception occurs. I'm not certain that this is a Weld problem, but it seem slike a good place to start based on the packages shown.</p> SEAMJMS-16

Event to JMS to Event mapping behavior

Bug Major Closed Rejected Unassigned John Ament Thu, 3 Feb 2011 21:56:15 -0500 Thu, 24 Feb 2011 11:32:56 -0500 Thu, 24 Feb 2011 11:32:43 -0500 0 3 <p>Looks like something doesn't properly switch TCCL context.<br/> e.g. your current top level CL is from jboss-service.xml, which is what is the content of JBOSS_HOME/common/lib<br/> Where it should really point to your app's CL.</p> <p>At least in what's currently in there, all components are in the same class loader/deployment, so a classloader problem would be a really bizarre one.</p> <p>OK, I guess you're not bundling HornetQ, which is where all of this starts.<br/> It looks like org.jboss.seam.jms.bridge.IngressMessageListener.onMessage should switch the TCCL to app's CL.</p> <p>Skywalker:lib alesj$ pwd<br/> /Users/alesj/jboss/build/target/jboss-6.1.0-SNAPSHOT/common/lib<br/> Skywalker:lib alesj$ ls | grep hornetq<br/> hornetq-bootstrap.jar<br/> hornetq-core.jar<br/> hornetq-jboss-as-integration.jar<br/> hornetq-jms.jar<br/> hornetq-logging.jar<br/> jboss-as-hornetq-int.jar<br/> Skywalker:lib alesj$ </p> <p>Hence jboss-service.xml's CL is used &#8211; this is where HornetQ comes in.</p> <p>HornetQ's not embedded in my deployment. It's deployed as a war file. These are my only libraries:</p> <p>jboss-logging-3.0.0.Beta4.jar seam-solder-api-3.0.0.Beta1.jar<br/> seam-jms-api-3.0.0-SNAPSHOT.jar slf4j-api-1.5.10.jar<br/> seam-jms-impl-3.0.0-SNAPSHOT.jar</p> <p>in case it helps, this is the code that creates the listener:</p> <p> private void createListener(final AfterBeanDiscovery abd, final BeanManager beanManager, final Route route, ConnectionFactory cf) {<br/> for(Destination d : route.getDestinations()) {<br/> try</p> { Connection conn = cf.createConnection(); Session session = conn.createSession(false, Session.AUTO_ACKNOWLEDGE); MessageConsumer cons = session.createConsumer(d); cons.setMessageListener(new IngressMessageListener(beanManager, route)); conn.start(); } <p> catch (JMSException e) </p> { log.error("Error creating listener",e); } <p> }<br/> }</p> <p>Rejecting issue. This is apparently as defined by spec, so cannot be fixed.</p> <p>Reopening, I don't understand what is spec required about this? Also, this is certainly fixable in JBoss AS, even if we have to do something complex to get there.</p> <p>Realized last night (with help from Stuart Douglas) that this issue (with TCCL) was being cauesd by hornetq sending messages outside of the application. By passing in the context class loader from another bean, we can set the class loader within the on message method, then set it back in a finally block. This allows the body of the on message to be executed within the app's class loader, fixing any issues with bean resolution.</p> <p>John, I obviously failed to explain this well to you, as this is what I was getting at when we were chatting ;- Sorry about that!</p> <p>John, I obviously failed to explain this well to you, as this is what I was getting at when we were chatting ;- Sorry about that!</p> <p>Rejecting, as this was a development issue that was believed by the dev to be a weld problem but is really a CL issue.</p> Date of First Response Fri, 4 Feb 2011 05:22:28 -0500 Global Rank 71680 Number of attachments 0.0 Number of comments 11.0 Rank 0|1094t4: Rank (Obsolete) 53392 Time in Status