While working on the exercise, a few questions arose:
I would have splitted the process elsewhere, as shown in the walktrough
The dispatcher could update the datatable and add the coustomer information.
So the performer could start directly with System3.
Wouldn’t it be better to have two performer and split the process again for more robustness?
One Performer for the Task in System3 (which adds the result to an orchestrator Queue) and another one for the final entries in System1?