A trigger is what starts a flow—it is the entry point. Each flow requires exactly one trigger, which starts the flow when the criteria for the trigger is met.

There are different types of flow triggers:

Cache Event Trigger

Cache Event triggers a flow as soon as there is a Create, Update, or Delete of the connector's data object selected in the trigger. 

Trigger fields showing the cache event option.

Completed actions also cache write data. In situations where data is maintained bidirectionally, it is important to set the Event Origin to indicate where the data was changed. Select the name of the connector to indicate changes made directly to the data from the external system. Select Ryvit to indicate changes made to the data via an internal action from the App Xchange platform. This will prevent you from having loops where you update one system and then immediately update the other system. 

Action Close Out Trigger

Action Close Out triggers a flow on the success or failure of an action.

Trigger fields showing the action close out option.

It is common to want to send a close out to the source system on the success of an action or remediate the data so an action can be retried on failure.

On Demand Trigger

On Demand triggers may or may not have a custom input of data. They can be scheduled to run at an interval of your choosing, or they may be defined as called flows and run from inside another flow.

On Demand triggers and called flows can simplify flow writing and make flows more reusable.

For example, imagine an integration that sends many different types of data between two systems, and one of the systems requires each of these data types to respond with a message on success or failure.

In this scenario, writing one called flow would allow you to then call that flow in each of the other data syncs. This would reduce the number of steps in each flow and allow you to manage any issues with that close out logic in one place.

Work Request Batch Ready Trigger

Use the Work Request Batch Ready trigger when writing flows that combine work request cache events into a single batch.

Flow authors can set a batch size limit for work requests, and when the limit is reached, it will create an event which will trigger the associated flow run.