I’m trying to model a business process for ordering (goods or service) and I really have difficulties understanding the correct use of events.
I have the following questions:
Is it correct to use the error events on the boundary of the subprocesses to express that in case of quality issues (“error”) the reclamation subprocess should be started?
The result event of the reclamation subprocess could be either one out of four. Is it correct to use the multiple event gateway to route the token coming from the reclamation process?
If I put everything inside the dashed line in a subprocess, do I have to add any other activity/event to the high-level process to stop it in case of a reclamation leading to the order cancellation event or the “accept with defects” event?
Any feedback beyond my questions, helping me to improve is welcome
Thanks in advance!
Welcome to the forum.
I’m not so sure, what kind of modeling tool you’re using, but if you where to use Cawemo or the Camunda modeler it would automatically do a lot validation for you that would prevent you from combining symbols that aren’t supposed to be combined.
But to answer your questions:
Sure, you can do this. The assumption then is that the sub process throws an error event that is caught by the boundary event.
They symbol you’re using in the model is an event based gateway but it’s used incorrectly in this case. firstly and event based gateway can only waiting for catching events that will occur some time in the future, not events that have occured in the previous process. Also the gateway will not send tokens through more than one sequence flow. I think what you should use is actually an OR gateway.
Putting everything in the dotted line has no affect at all. It’s a group just is just a visual indicator and has no impact on the process beyond visual ascetic.