Activity looping/parallel for each: Output mapping and Inturrupting
we want to model and execute an approval process with one requestor and several approvers. All assigned approvers get the approval task simultaniously to save time. Each approver can make a decision like Reject, Rework, Approve and also add comments.
Only if all Assignees make their approval, the request gets approved. On the first occuring "rework" or "reject" decision, all open tasks have to be removed from the assignees UWLs; their results become unnnecessary.
Please have a look at the corresponding model on [http://www.gm.fh-koeln.de/~wi1744/pubfiles/parallel_for_each_task.PNG].
As you can see in the model, we made an approach using a parallel for each Human activity. In our process the tasks show up in parallel on each adressed UWL. Thats fine so far but we are facing two problems.
1. Output Mapping of parallel Tasks
1.1 Are we right to assume that for each task output the activitys output mapping is done?
1.2 Are there differences to the regular output mapping we have to consider?
2. Stopping the parallel tasks and continuing the process flow .
We said as soon as any result like rework or reject comes out, the whole activity should stop and all open tasks be removed from the UWL. How can this be modeled and realised?
Thank you for any input!