Troubleshooting Your Flow
A flow might fail for a variety of reasons. Understanding the various types of error messages can help you isolate the problem and fix it.
View error messages in the Runs tab. In general, errors can be grouped into the following categories:
Step input validation
Script evaluation
Downstream dependency failure
Step Input Validation
This type of error occurs when the input is invalid for the particular flow step.
The input of a flow step is dynamic, so you won’t always know what the input is going to be until you run the step. A step can fail if it receives incorrect input values, and it will produce an error message indicating what aspect of the input is invalid.
For example, a Combine Delimited Files step may attempt to combine files, but it will fail if the input does not point to any files.
Script Evaluation
This type of error occurs when a flow step has an issue evaluating and executing a script.
Scripts do the actual data transformations, and they power many aspects of a flow run. This is because the main purpose of a flow is to take in data and transform it so that it can integrate into a target system. Therefore, script execution errors are common while setting up a flow.
For example, this is a script error that is the result of trying to evaluate a script as an input to a map step:
This is because the input script is an invalid format:
Another type of script evaluation error occurs when the script being executed has an error in it. These are more difficult to troubleshoot, as you first need to understand what is wrong with the script before you can attempt to fix it.
For example:
In this case, the flow step tried to access a property (length) on an undefined object. The script looked like this:
You can determine that the undefined object is the property records, as that is where the length property is supposed to pull data from. This means that the flow’s trigger data does not contain the property records. Validate that by looking at the flow’s trigger data in the Run Details JSON:
This shows there is no data in the flow’s trigger, which is what is causing the undefined object error in this case.
To fix this issue, you need to determine if it is normal in this instance for the trigger data to be empty.
If no data is expected, then you can fix the script using JavaScript falsy logic. For example, input something such as:
If the flow’s trigger should contain data, then you should stop the flow with a Stop Flow step and appropriate messaging that describes the issue. For example, add a Conditional step with a nested Stop step to stop the flow if the trigger data contains no records:
This would set the flow up to display the following run result:
Downstream Dependency Failure
This type of error comes up when there is an issue in a flow dependency, not the flow itself.
Flow steps often interface with APIs, so it is possible for a failure to happen at the flow step level due to that dependent API failing. In general, this type of error is not permanent and will resolve if the flow is run a second time.
This is why it is important to follow Best Practices for Flow Design and test your flows to make sure they are robust and execute with expected results. If a flow encounters a downstream dependency error, the flow must be able to run again and not duplicate data. If applicable, you can also improve the resiliency of your flow by enabling a retry policy for interactions with external systems.
To troubleshoot downstream dependency errors, look at the error message and determine if the flow should be re-run. All interactions with dependencies should have retry policies in place, but it is possible to see a dependency error even with those policies. The error message will indicate what status code the dependency returned, and you can take action based on that.