From what I recall, once data has been written to the TN database, its basically read-only, except for changing a few bizdoc attributes. The only major you can do is to delete the bizdoc completely. I think this is because of the performance hit of persisting XML. This approach also fits in with TN’s peculiar niche - initiating processing carried out by the IS document translation engine beneath it, and logging the document transformation stages.
What we do here is to just add new content parts. Since the content part name must be unique, we just have the timestamp (down to the millisecond) as part of the content part name - this works pretty well.
#webMethods#Integration-Server-and-ESB#B2B-Integration