Adding and Configuring an Event Trigger
Pipeline triggers are used to initiate a pipeline event and to start data flowing down a pipeline. The event trigger is unique in that it does not initiate events on a specific schedule but can watch certain types of data sources for real-time updates, instantly delivering new or modified data as soon as a change occurs.
The event trigger can be added to a pipeline from the Stages > Trigger > Event menu at the top of the pipeline builder interface. Once added, the input generating the event can be selected from the right-side references panel.
Event Types Supported by the Event Trigger
Event Triggers can only monitor certain types of inputs.
The types of connection inputs that event triggers can generate events from are as follows:
Special Considerations of an Event Trigger
The configuration of the event trigger will allow non-eventable inputs to be entered as the reference, but they will not generate events when changes are made to those data sources. Only the data source inputs listed above will function properly in event triggers.
Some high-traffic data systems can generate or change data frequently. In these cases, it should be ensured that any pipeline can complete a run before the next event starts on average – some queuing is acceptable in burst-like workloads. Event triggers and pipelines each have inherent queues for up to 10,000 messages. For data with high-frequency updates, it may be more effective to use a polled trigger to sub-sample data, especially when individual data points are less critical than capturing consistent trends over time.