Hello,

We have a TwoWay receive port where the schema is exposed as a Web service.

We receive XML wrapped HL7 ORM (LabOrder) message on the receive location which has a custom pipeline, in custom pipeline we have two custom pipeline components 1. Logging Component (which logs the raw message to sql database) 2) Custom HL7XMLDisassemler.

We want to send NACK from within the Logging pipeline component back to the client when exception occures. For this purpose I am catching the Sql exception and creating a new IBaseMessage and promoting the properties like RouteDirectToTP, EpmRRCorrelationToken, ReceivePipelineResponseConfig, CorrelationToken, ReqRespTransmitPipelineID and IsRequestResponse. (These properties are getting promoted in NACK out message, verified debugging the pipeline component)

Even promoting this properties the message gets consumed by next ppeline component and client does not ger NACK.

The same works in case of ORU messages which we receive on MLLP adapter. We do get NACK back if sql server is down or not reachable.

Please suggest if  am missing any steps or it differs in case of MLLP and WCF request response.

Thanks a lot...

• Edited by Dhiraj Bhavsar Monday, May 25, 2015 7:36 AM

Even promoting this properties the message gets consumed by next ppeline component and client does not ger NACK.

Any other explanation/hints about this phenomenon?