We are forcing the FTP get into a content stream by setting the threshold at 0. the result, is that the data is seen as a contenet stream, and the representing data populates the tspace location as text. We read part of the stream (mark and reset) to determine if the header is unique, and exit or put accordingly.
Both the exit and put close the stream after they are done. However, the tspace location still has a file sitting there. It almost appears, that the IS thinks they are still attached or in use, because you cant delete them until after the IS has been re-started.
we do have TN on this cluster, but we are not calling any of the TN code.
I am curious if GC will see the data in the tspace location and clean it out, or if it only removes trash from memory.
#webmethods-Protocol-and-Transport#Integration-Server-and-ESB#webMethods