I'm using a Wrapped Metanode that runs well in default mode, but when switched to Simple Streaming, throws the following exception:
2016-07-22 15:06:33,798 : ERROR : Streaming-203-Master : SimpleStreamerNodeExecutionJob : Analysis : 4:213 : Caught "NullPointerException": Argument must not be null. 2016-07-22 15:06:33,798 : DEBUG : Streaming-203-Master : SimpleStreamerNodeExecutionJob : Analysis : 4:213 : Caught "NullPointerException": Argument must not be null. java.lang.NullPointerException: Argument must not be null. at org.knime.core.data.filestore.internal.ReferenceWriteFileStoreHandler.<init>(ReferenceWriteFileStoreHandler.java:72) at org.knime.core.node.workflow.NativeNodeContainer.initFileStore(NativeNodeContainer.java:717) at org.knime.core.node.workflow.NativeNodeContainer.mimicRemotePreExecute(NativeNodeContainer.java:455) at org.knime.core.node.workflow.WorkflowManager.mimicRemotePreExecute(WorkflowManager.java:4349) at org.knime.core.node.workflow.SubNodeContainer.lambda$3(SubNodeContainer.java:972) at org.knime.core.node.workflow.SubNodeContainer.runParentAction(SubNodeContainer.java:1223) at org.knime.core.node.workflow.SubNodeContainer.runIfInExternalExecutor(SubNodeContainer.java:960) at org.knime.core.node.workflow.SubNodeContainer.performStateTransitionPREEXECUTE(SubNodeContainer.java:972) at org.knime.core.node.workflow.WorkflowManager.doBeforePreExecution(WorkflowManager.java:2877) at org.knime.core.node.workflow.NodeContainer.notifyParentPreExecuteStart(NodeContainer.java:499) at org.knime.core.node.workflow.NodeExecutionJob.internalRun(NodeExecutionJob.java:136) at org.knime.core.node.workflow.NodeExecutionJob.run(NodeExecutionJob.java:110) at org.knime.core.streaming.SimpleStreamerNodeExecutionJobManager$1.run(SimpleStreamerNodeExecutionJobManager.java:94) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) at java.lang.Thread.run(Thread.java:745)
Inside my node, there are some other metanodes, some Image Reader nodes, a few more nodes from the Image Processing extension, and a bunch of GroupBy, Joiner, and Pivot nodes (that are not streamable but to my knowledge shouldn't block the execution in streaming mode).
Are there some rules on what nodes I can put inside a streamed metanode, or some known things to avoid?
Thanks,
Jan