1) If everything goes normal, after Task One is executed, and the called-process triggered by Call Activity is completed, user arrives at Task Two.
2) If after executing Task One, called-process triggered by Call Activity is not finished, and the boundary signal is thrown, then user arrives to Task Two without waiting for the called-process to finish.
My problem is, that during scenario 2, if called-process completes at some point after main process is already completed, user still arrives at an active Task Two, even if "cancelActivity = true" for the boundary event.
Is this proper behavior? If yes, is adding additional check upon e.g. process variable a practical way to implement the process?
Hi, I see you have a processEnded = true condition, I guess you want to say "processEndedAfterSignalThrown". yes, it seems to be an inventive solution for what are you trying to do
I've played a bit more with process diagram, and came to this conclusion: the first scheme is wrong in any case. I want the main process to be properly completed regardless of the path. However, if I end call activity with a boundary event, the process is stuck waiting in the second parallel gateway forever, — is my understanding correct?
not sure what is the behaviour (never tested it) but if i have to guess probably will wait the subprocess to be canceled or completed But a unit test to show the behaviour would be welcome.