Long Delay in CommandContextInterceptor: context.close()

[j2eeman84]

I have a subprocess (called activity) that consists of a parallel gateway that connects to
two exclusive gateways.

Based on some logic in the called activity's process listener, a single task is created or skipped after the top
exclusive gateway and has a comma separated list of candidate users.

Based on some logic in the called activity's process listener, a multi instance task is created or skipped after
the bottom exclusive gateway.

          1) Child (Called) Activity
              LINK: https://app.box.com/s/ysw72iybepuweydbwp5q

          2) Parent (Set of multi instance tasks that only call the above activity)
              LINK: https://app.box.com/s/cbe09uedied6ajacszm7


The issue that I am having is that every time a top level multi instance call is made to the called activity process
and the process listener completes. The creation of my called activity's tasks are halted for atleast 20-50 seconds.

After doing remote debug, I am finding that the halt occurs at the CommandContextInterceptor: context.close()
method. Before the final flush occurs. I am using an Oracle DB and Activiti 5.16.4.
I have the jobexecutor turned off and my tasks are all exclusive.

Any clue what the extreme hold up is? The application is a candidate for deployment at a large University if
I can get this to work properly. As far as functionality, everything works correctly.

Thanks,

[trademak]

Is it possible to write a unit test that shows the issue?

Best regards,