Flow Steps

Flow steps define specific functions that accomplish a variety of tasks. Steps carry out the actual work of a flow, and there are a variety of different types of flow steps. Steps can perform independent functions, or they can work with other steps.

For example, if you want to confirm that an RFI exists, you could use a Lookup step to check that an RFI ID exists. If you also want to confirm that the RFI ID matches a given range of IDs before proceeding with the flow, you could pass the output of the Lookup step to a Conditional step, where you compare against an array of expected IDs.

To add flow steps to your flow, see Add Flow Steps. For details about each type of flow step, review the Flow Step Index.

Processing Flow Data

Flow data is processed in the App Xchange flow engine. The step drawer UI (the Edit Step window) is a visual tool for writing the JSON that is passed into the flow engine. 

The flow engine processes the flow and its steps. All steps have access to the flow’s trigger data, event info, and any steps that have previously run.

Some steps are more complex than others. For example, in a Parse CSV File to JSON Step, you supply it with the file contents. However, in a Post to a Network App Step, you will need to write more code.

Validate that a Flow is Working

Flow writing is a lot like programming. Start small and test as you go. In most situations, you can run the flow and check the flow run results to verify that the step input and output is what you expect. For more information, see Test a Flow.

When steps fail, the entire flow will typically fail, and you’ll see an error in the flow run result. For more information about what to do when a flow fails, see Troubleshoot a Flow. For more information about flow run results, see Flow Run History.