e.g. if i have an action associated with a task node where i read an important process variable, and for some reason cannot get a proper value,
the workflow should come to an halt so an admin guy can sort the technical problems.
this is just an example, it could be anything;
in a handler I use a connection to another system and it is not working,
a connection to an EJB is not working,
At the moment I have created a special "Admin" node, which the workflow is redirected to in case of technical problems, and the admin can fix the
technical problem and assign the task back when it is sorted.
The problem with that solution is that the "Admin" node is shown in the workflow xml, which makes the workflow diagram very messy
(since every node that cause problems must have a transition to the "Admin" node).
How do you guys handle technical problems in the workflow?
You can try creating a superstate, adding the nodes that cause problem and pulling the transition from the superstate (so you can take it from any node).
A second approach is an exception handler at the process level. However, you won't be able to define a node there.