Handle Node Failure
The On Fail feature allows users to specify how individual nodes behave in case of logic failure.
By defining On Fail behaviors, users can effectively manage workflow execution, ensuring seamless handling of node errors.
Enabling On Fail Behavior
To enable the On Fail for a node, open the Node Menu (•••) for the desired node.
Click on "Add On Fail" to append an On Fail behavior section to the node. Users can pick from three On Fail behaviors:
- Retry
- Terminate
- Custom Flow
Understanding On Fail Behaviors
Retry
This behavior allows the system to automatically attempt the node logic again within specified intervals. Users can set the maximum number of retries and determine if the workflow should continue after retry attempts.
| Configuration | Purpose |
|---|---|
| On Fail | Retry |
| Max Retries | Define the maximum number of retry attempts the backend should make. |
| Interval (ms) | Set the time interval between retries (maximum: 10000ms). |
| Continue after retry | Determine whether the workflow should proceed after exhausting the retry attempts. |
Terminate
This option terminates the entire workflow if the specific node fails, ensuring an immediate halt to the execution.
| Configuration | Purpose |
|---|---|
| On Fail | Terminate |
Custom Flow
Users can create a custom sub-workflow to execute if a node fails, offering a tailored response to node failure. By editing the sub-flow, users can add nodes from the library, paste from clipboard, or generate nodes using AI.
| Configuration | Purpose |
|---|---|
| On Fail | Custom Flow |
| Custom Flow | Define a sub-workflow to execute if the node fails. |
| Continue After Custom Flow | Choose whether the main workflow continues after executing the sub-workflow. |