Node Error Handling Strategy

1. What is the Error Handling Strategy

The Error Handling Strategy is used to set how to handle node execution failures.

During Workflow execution, nodes may fail due to request failures, insufficient permissions, missing documents, field extraction failures, configuration errors, and other reasons. Through the Error Handling Strategy, you can decide how to proceed after a node error, avoiding unnecessary interruption of the entire Workflow.

Different nodes may support different error handling methods; please refer to the options displayed in the node configuration panel for details.

2. Common Error Handling Methods

Most nodes that support error handling offer the following three options:

• No Handling: When the node execution fails, the Workflow execution will be interrupted.
• Default Value: When the node execution fails, the Workflow continues to run, using a pre-configured default value as the node's output.
• Error Handling Branch: When the node execution fails, the current path is interrupted, and the flow switches to the connected error handling branch.

3. Nodes Supporting Error Handling

Error handling is currently available for the following nodes:

Different nodes may support different error handling methods; please refer to the options displayed in the node configuration panel for details.

4. Default Value Handling

If error handling is set to Default Value, the workflow doesn't stop when a node fails. Instead, it continues running and substitutes the configured default value for that node's output.

The default value configuration items are generated based on the node's original output variables. In other words, the default value configuration usually corresponds to the variables that the node would have originally output.

For example:

• If the node originally outputs Text, then Text will appear in the default value configuration.
• If the node originally outputs Result, then Result will appear in the default value configuration.
• If the node's output variables are user-defined, the default value configuration will be generated based on the output variables configured by the user.
• If it is a Tool or App, the default value configuration will be generated based on the output variables originally defined by that Tool or App.

If some default values are not required, leaving them blank usually outputs an empty string or ignores that output variable. Please follow the page prompts for specifics.

5. Default Value Applicability Scope

The following nodes provide a Default Value Applicability Scope configuration when Default Value is selected:

• Document Read node
• JSON Variable Extractor node

This is because these two nodes may process multiple output items in one execution and may encounter "partial success, partial failure" situations.

For example, the Document Read node can read multiple documents in the same node. If a document path is incorrect or the document does not exist, only the reading of that document may fail, while others can still be read normally.

The JSON Variable Extractor node may also encounter some fields extracted successfully and others failed when extracting multiple fields.

In this case, the Default Value Applicability Scope can control how the default value is applied:

• Only Error Items: Only the failed items use default values, while the successful items retain the actual results.
• All Items: If any item fails, all items use their respective configured default values.

If you want to preserve successful results as much as possible, it is recommended to select Only Error Items.

6. Special Error Handling for Document Write and Document Delete

Operation-type nodes such as Document Write and Document Delete differ from most nodes: they usually do not use "default values" to replace outputs, but instead control error handling after node failure through Ignore Errors.

When Ignore Errors is selected, the Workflow continues running after the node execution fails.

If multiple documents are configured in the node, you can also set the Ignore Scope:

• Only Error Documents: Only ignore the documents where the operation failed; other documents will still be processed normally.
• Entire Node: When any document operation fails, errors are ignored for the entire node.

This mechanism is suitable for scenarios where multiple documents are written to or deleted in one node, reducing the impact of a single document failure on the overall process.

7. Error Handling for Loop Nodes

Loop nodes have independent error response methods.

Since loop nodes typically process multiple elements, there may be situations where one element fails to process, but other elements can continue processing.

Loop nodes support two error response methods:

• Terminate on Error: Terminate the loop node and output error information when an exception occurs.
• Remove Error Output: Ignore the exceptional elements, continue processing the remaining elements, and only retain normal results in the final output.

If you do not want a single element exception to affect the entire loop, you can choose Remove Error Output.

8. Nodes Not Supporting Error Handling

Error handling is currently not available for the following nodes:

These nodes will not show an "Error Handling" configuration item.

9. Usage Recommendations

• If the node should not continue executing after an error, select No Handling.
• If you want the Workflow to continue running and provide a fallback output, select Default Value.
• If you want to handle errors separately, such as sending alerts, logging, or executing backup flows, select Error Handling Branch.
• If it is an operation-type node such as Document Write or Document Delete, and you want partial failures not to affect subsequent processes, you can select Ignore Errors.
• If it is a loop node, and you do not want a single element exception to affect the overall result, you can select Remove Error Output.