We are new to V2R6.2 and TPT. Right now, we are only using it in the context of IBM's DataStage software - its TD Connector uses the TPT to interact with Teradata. We have many FastLoads, FastExports, MultiLoads and TPumps that will eventually be migrated to TPT as well.
I am trying to understand the nature of TPT vs the standalone utilities. Let's use FastLoad and TPT's load operator for the sake of conversation. I have had one person tell me that the TPT operator is calling all new code using new API's and everything, and there is no code shared with FastLoad. Another told me they both work with the same set of API's, and yet another told me that TPT's load operator actually FastLoad under the covers.
How does it actually work?
I apologize if this has been asked before, or if it is discussed in the documentation. If that is the case, please let me know and send me the links. I did a cursory search before posting, but not an exhaustive search either online or in the documentation.
The client-side code for the operators and the TPT framework / API is new, but the LOAD operator still uses the "FastLoad protocol" to communicate with the DBMS. So each instance of a LOAD operator is equivalent to a FastLoad as far as the database is concerned. Same applies for UPDATE operator / MultiLoad, or EXPORT operator / FastExport.
Most other operators use normal SQL session protocol, as do BTEQ and TPump.