[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld commented on JBRULES-3209: ----------------------------------------- Actually, what's causing this is that before september 13th, when the session was marshalled, no initialFactNode was retrieved and marshalled -- and now that _is_ happening. {code} public static void writeInitialFactHandleRightTuples(MarshallerWriteContext context) throws IOException { ObjectOutputStream stream = context.stream; InternalRuleBase ruleBase = context.ruleBase; ObjectTypeNode initialFactNode = ruleBase.getRete().getEntryPointNode( EntryPoint.DEFAULT ).getObjectTypeNodes().get( ClassObjectType.InitialFact_ObjectType ); // do we write the fact to the objecttypenode memory if ( initialFactNode != null ) { // <-- this if clause was not entered before! {code} -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld commented on JBRULES-3209: ----------------------------------------- The following change caused marshalling to break: https://github.com/droolsjbpm/drools/commit/0a3575e8073f2abc035b13e113814... {code} - public static final ClassObjectType InitialFact_ObjectType = new ClassObjectType( InitialFact.class ); + public static final ClassObjectType InitialFact_ObjectType = new ClassObjectType( InitialFactImpl.class ); {code} It looks like this results in the initialFact being marshalled whereas the old code didn't cause that. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld updated JBRULES-3209: ------------------------------------ Priority: Major (was: Minor) -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld updated JBRULES-3209: ------------------------------------ Summary: Unmarshalling a marshalled session fails when a fact has been inserted into the session (was: ReloadSessionTest fails when trying to unmarshall FactHandle) Description: Unmarshalling a marshalled session fails when a fact has been inserted into the session. This can be seen in the org.drools.persistence.session.ReloadSessionTest, which fails at the moment. An npe is thrown here: {code} ConcurrentNodeMemories.getNodeMemory(NodeMemory) line: 65 ReteooStatefulSession(AbstractWorkingMemory).getNodeMemory(NodeMemory) line: 1040 InputMarshaller.readFactHandles(MarshallerReaderContext) line: 373 InputMarshaller.readSession(MarshallerReaderContext, int, ExecutorService, Environment, SessionConfiguration) line: 255 DefaultMarshaller.unmarshall(InputStream, KnowledgeSessionConfiguration, Environment) line: 92 SessionMarshallingHelper.loadSnapshot(byte[], StatefulKnowledgeSession) line: 91 SingleSessionCommandService.initKsession(Integer, KnowledgeBase, KnowledgeSessionConfiguration) line: 221 {code} was: Unmarshalling a marshalled session fails when a Fact has been inserted into the session. This can be seen in the org.drools.persistence.session.ReloadSessionTest, which fails at the moment. An npe is thrown here: {code} ConcurrentNodeMemories.getNodeMemory(NodeMemory) line: 65 ReteooStatefulSession(AbstractWorkingMemory).getNodeMemory(NodeMemory) line: 1040 InputMarshaller.readFactHandles(MarshallerReaderContext) line: 373 InputMarshaller.readSession(MarshallerReaderContext, int, ExecutorService, Environment, SessionConfiguration) line: 255 DefaultMarshaller.unmarshall(InputStream, KnowledgeSessionConfiguration, Environment) line: 92 SessionMarshallingHelper.loadSnapshot(byte[], StatefulKnowledgeSession) line: 91 SingleSessionCommandService.initKsession(Integer, KnowledgeBase, KnowledgeSessionConfiguration) line: 221 {code} -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld commented on JBRULES-3209: ----------------------------------------- Ported the test to 5.2.x -- the same problem exists there. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld updated JBRULES-3209: ------------------------------------ Assignee: Marco Rietveld (was: Mark Proctor) Priority: Minor (was: Major) Description: Unmarshalling a marshalled session fails when a fact has been inserted into the session. This can be seen in the org.drools.persistence.session.ReloadSessionTest, which fails at the moment. An npe is thrown here: {code} ConcurrentNodeMemories.getNodeMemory(NodeMemory) line: 65 ReteooStatefulSession(AbstractWorkingMemory).getNodeMemory(NodeMemory) line: 1040 InputMarshaller.readFactHandles(MarshallerReaderContext) line: 373 InputMarshaller.readSession(MarshallerReaderContext, int, ExecutorService, Environment, SessionConfiguration) line: 255 DefaultMarshaller.unmarshall(InputStream, KnowledgeSessionConfiguration, Environment) line: 92 SessionMarshallingHelper.loadSnapshot(byte[], StatefulKnowledgeSession) line: 91 SingleSessionCommandService.initKsession(Integer, KnowledgeBase, KnowledgeSessionConfiguration) line: 221 {code} This _only_ happens when the knowledge base is reinitialized before reloading the statefulKnowledgeSession: {code} KnowledgeBuilder kbuilder = KnowledgeBuilderFactory.newKnowledgeBuilder(); kbuilder.add( ResourceFactory.newByteArrayResource(rule.getBytes()), ResourceType.DRL); if (kbuilder.hasErrors()) { fail(kbuilder.getErrors().toString()); } KnowledgeBase newkbase = KnowledgeBaseFactory.newKnowledgeBase(); kbase.addKnowledgePackages(kbuilder.getKnowledgePackages()); StatefulKnowledgeSession newCommandKSession = JPAKnowledgeService.loadStatefulKnowledgeSession(sessionInfoId, newkbase, null, env); {code} was: Unmarshalling a marshalled session fails when a fact has been inserted into the session. This can be seen in the org.drools.persistence.session.ReloadSessionTest, which fails at the moment. An npe is thrown here: {code} ConcurrentNodeMemories.getNodeMemory(NodeMemory) line: 65 ReteooStatefulSession(AbstractWorkingMemory).getNodeMemory(NodeMemory) line: 1040 InputMarshaller.readFactHandles(MarshallerReaderContext) line: 373 InputMarshaller.readSession(MarshallerReaderContext, int, ExecutorService, Environment, SessionConfiguration) line: 255 DefaultMarshaller.unmarshall(InputStream, KnowledgeSessionConfiguration, Environment) line: 92 SessionMarshallingHelper.loadSnapshot(byte[], StatefulKnowledgeSession) line: 91 SingleSessionCommandService.initKsession(Integer, KnowledgeBase, KnowledgeSessionConfiguration) line: 221 {code} -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/JBRULES-3209?page=com.atlassian.jira.plug... ] Marco Rietveld updated JBRULES-3209: ------------------------------------ Comment: was deleted (was: I'm leaving Friday (tomorrow), so I'm assigning it to Mark, who will be able to fix it far more quickly than I can. ) -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira