Skip to main content
Version: 8.7

Ad-hoc sub-processes

Ad-hoc sub-processes are a special kind of embedded subprocesses with an ad-hoc marker (represented by a ~ tilde character). Compared to regular subprocesses, ad-hoc sub-processes allow more flexibility for executing inner elements.

A demo process with an ad-hoc sub-process. Some elements inside the subprocess are active.

The inner elements of an ad-hoc sub-process are not connected to a start or end event. Each element can be executed multiple times, in any order, or skipped.

If elements depend on each other, the elements can be connected by a sequence flow to build a structured sequence within the ad-hoc sub-process.

When a process instance reaches an ad-hoc sub-process, it activates the inner elements and waits for their completion. After the last element is completed, the process instance completes the ad-hoc sub-process and takes the outgoing sequence flows.

Constraints​

An ad-hoc sub-process has the following constraints:

  • Must have at least one activity
  • Must not have start events or end events
  • Any intermediate catch event must have an outgoing sequence flow

Activate an element​

An ad-hoc sub-process can define an expression activeElementsCollection that should return a list of strings. Each string in the list should match to an ID of an inner element of the ad-hoc sub-process. Usually, the expression accesses a process instance variable that was created before and holds the list of element IDs.

A process with an ad-hoc sub-process that shows how a variable is used to active the inner elements.

When a process instance reaches an ad-hoc sub-process, it evaluates the expression activeElementsCollection and activates all elements whose element IDs are in the list.

If the list is empty or the expression is not defined, no element is activated and the ad-hoc sub-process remains active.

If the expression doesn't evaluate to a list of strings, or the list contains other values than inner element IDs, the process instance creates an incident.

note

Currently, it is not possible to activate elements dynamically after the ad-hoc sub-process is activated, only on entering the subprocess.

Variable mappings​

An ad-hoc sub-process can define input and output variable mappings.

Input variable mappings are applied on activating the ad-hoc sub-process and before evaluating the expression activeElementsCollection. They can be used to create local variables for the ad-hoc sub-process.

Output variable mappings are applied on completing the ad-hoc sub-process. They can be used to propagate local variables from the ad-hoc sub-process into the process instance. By default, no local variables are propagated.

Additional resources​

XML representation​

<bpmn:adHocSubProcess id="ad-hoc-subprocess" name="Ad-hoc sub-process">
  <bpmn:extensionElements>
    <zeebe:adHoc activeElementsCollection="=activeElements" />
  </bpmn:extensionElements>
  ... more contained elements ...
</bpmn:adHocSubProcess>