How Duplicate Detection Works
When you enable the duplicate detection feature, as described in
Configuring
Process Engine Settings and
Requirements for Using the Duplicate
Event Detection Feature,
webMethods Broker (deprecated),
Integration Server, and
Process Engine work together to detect duplicate documents to prevent the creation of a redundant process for a duplicate document.
Initially, webMethods Broker (deprecated) detects when a document is a duplicate. To enable duplicate detection, you modify the generated triggers to identify the wm.prt.duplicate:CheckDuplicates service as the service to resolve duplicate documents. As a result, when webMethods Broker (deprecated) detects a duplicate document, it invokes the wm.prt.duplicate:checkDuplicates service.
The wm.prt.duplicate:checkDuplicates service determines how to handle the duplicate document. If the service determines that:
Process Engine has not yet received the document—the service returns the NEW status, which in turn, causes the
Integration Server to deliver the document as usual.
Process Engine has previously received the document and the Process Engine has already completed processing the document—the service returns the DUPLICATE status, which in turn, causes the
Integration Server to dispose of the duplicate document.
Process Engine has previously received the document and the Process Engine has started processing the document, but processing is not complete—the service returns the IN DOUBT status, which in turn, causes the
Integration Server to write an entry to the WMDOCUMENT table of the Process Audit database component so that a
webMethods Monitor user can decide how to handle the document.