The usual reason that a TPA isn’t used is due to identifier mismatches–the TPA sender/receiver doesn’t match the sender/receiver in the received document. Double-check your identifiers in the profiles and in the document. That said, is there a reason you don’t want TN to do the splitting?
If you’re not going to use TN to do the splitting, then you have to do all the work yourself. The bizdoc of the X12 envelope has no idea what the transaction sets are–because it hasn’t split them to look at them. If you want to process at the envelope level, you’re going to need to split and determine the type of each transaction set yourself. Don’t reinvent the solution here–TN already has all you need to do the splitting and the transaction set-specific processing.
This var won’t be set for an X12 envelope document.
Why not? You’ll have multiple transaction sets, so it makes sense that the service that processes that type of transaction set will be called multiple times.
The best approach to this technical solution (this isn’t a business case) is to let TN do the splitting. Use the TPA settings to generate FAs. Have a rule to process transaction sets. You don’t need a rule for envelopes. You can have a rule to ignore groups. Let TN do the work it was designed for. Unless you have some very specialized processing needs, you shouldn’t need to process the envelope yourself.
#edi#Integration-Server-and-ESB#webMethods