Resolution:

After installing BE 2.2, Hotfix 1, the Secondary Inference Engine throws the following exception.

Exception
-------------
Failed to start the FTAsyncRuleServiceProviderImpl
java.lang.RuntimeException: Node Node[GUID=test.na.tibco.com,ID=3,Priority=2,State=NODE_CREATED] exists in the nodemap.
    at com.tibco.cep.runtime.service.om.tangosol.AbstractFTNodeManager.addNode(AbstractFTNodeManager.java:300)
    at com.tibco.cep.runtime.service.om.tangosol.AbstractFTNodeManager.start(AbstractFTNodeManager.java:134)
    at com.tibco.cep.runtime.session.impl.FTAsyncRuleServiceProviderImpl.run(FTAsyncRuleServiceProviderImpl.java:317)
    at java.lang.Thread.run(Thread.java:595)

Resolution

=======

Check if the same project works without the hotfix. If so then this exception is thrown due to a misconfiguration.

As per BE 2.2- HF 1 release notes:

Fault tolerance node names changed automatically upon recovery from situations such as network disconnections, causing system failure in fault tolerance groups. A new engine property enables you to specify unique and stable node names for each engine in the FT group. A node with the same name as an existing node cannot join the cluster. Use the new property "be.ft.nodename" to specify the name in each node's be-engine.TRA file. The node name cannot be more than 30 characters. If the node name property is not specified in a FT setup the engine name is used as the default node name which is specified by using "-name"  argument with be-engine executable in a standalone deployment or by TRA property "be.engine.name" in TIBCO administrator based deployment. The engine name which can be more than 30 characters needs to be overridden in this situation by the property "be.ft.nodename" .

After installing BE 2.2, Hotfix 1, the Seconday Inference engine throws the exception, "Failed to start the FTAsyncRuleServiceProviderImpljava.lang.RuntimeException: Node Node[GUID=test.na.tibco.com,ID=3,Priority=2,State=NODE_CREATED] exists in the nodemap".