Published on May 3, 2019
IBM App Connect Enterprise software has provided event-driven processing nodes for many years, to control the flow of messages through your message flows. These nodes provide message flow aggregation, message collection, message sequencing, and timeout capabilities. The Aggregation, Collector, Sequence, and Timeout nodes require a queue manager to be present on the same computer as the nodes themselves in order to persist the event state. For this reason, the nodes were not available in the App Connect enterprise capability on IBM Cloud because it doesn’t provide a local queue manager.
In the latest release of App Connect, this restriction has been removed, so integrations that contain event-driven nodes can be run. However, the nodes will not persist their state, so if the integration is restarted for any reason, the event state will be lost. For many use cases for these nodes, this limitation is acceptable if a long-lived persistent state isn’t needed. Typically, if event processing is expected to complete in a few minutes, the lack of persistence is unlikely to cause any issues.
No new steps are required to configure App Connect. When you upload a BAR file that contains one or more event-driven nodes and then start the associated integration server, the required queue manager (which uses non-persistent storage) is created for use by the nodes. The “Import a BAR file” panel provides some information about the limitations of the running nodes:
For full details about how to configure both App Connect (with a plan that provides enterprise capabilities) and event-driven nodes, see the tutorial Constructing an App Connect integration that uses aggregation, collection, resequencing, and timeout event-driven mechanisms.
If you’d like to see a persistent version of the event-driven nodes added to App Connect, please add a comment below with details about your use case.