]
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}
ReloadSessionTest fails when trying to unmarshall FactHandle
------------------------------------------------------------
Key: JBRULES-3209
URL:
https://issues.jboss.org/browse/JBRULES-3209
Project: Drools
Issue Type: Bug
Security Level: Public(Everyone can see)
Components: drools-core
Reporter: Marco Rietveld
Assignee: Marco Rietveld
Priority: Minor
The org.drools.persistence.session.ReloadSessionTest fails at the moment.
I'm seeing a npe 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}
The cause of this seems to be that when
InputMarshaller.readFactHandles(MarshallerReaderContext) runs in this test, the sinkId
retrieved from the stream is _6_, but should be 2. The context contains an ObjectTypeNode
with id 2, but not 6, in other words (and id 2 maps correctly).
When I change the value (during debugging) to 2, the test completes successfully.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: