Error Recovery

Recover processes in an error state and move them to the node of your choosing.

A process might error for many reasons. You might set character data in a numeric or date field. You might loose connectivity to your SMTP gateway. A variety of system and data specific conditions might cause a process to error. This node listens for process in error for a given workflow, redirects them to a specific node, and readies the process again.

This node serves as an improved replacement for the “Error Import” node. The legacy node has been deprecated. While it may still be available in the SDN, all users are encouraged to use this version of the node.

Configuration

• Set a Frequency for error checking that is appropriate for your process, or leave the default enabled at 15 minutes.

• Optionally specify the Title of the Error Node. The error node is the node where the error occurred. Leaving this field blank will result in ALL errors, regardless of where they happened, getting picked up. When setting a value for the error node, you should use multiple Error Recovery nodes in your workflow (1 for each error condition you seek to pick up). You can not name one error node and sweep all the rest with an second, empty error node.

• Set the Restart Node. This is the target node error processes will be restarted in. This is often an Validation node.

Please note, the Title specified in the Restart Node field controls the routing. The linked node (the node the error recovery node is connected to in the workflow design) is not relevant. More on this below.

Both the Error Node and Restart Node fields expect the user entered Title for the node. Each node has a title as seen here:

Workflow

The most common deployment for a workflow would be to have a single Error Recovery node routing all error conditions. The Error Recovery node can be linked to any node on the design surface, but the Restart Node is what governs where error processes are sent. For ease of understanding the workflow for future users, it might make sense to have the error recovery visually linked to the “Restart Node”.

Alternately, error import can also target an End node. Both are valid deployment options.