Problem: When for a node is defined the setback as error handling behaviour, you specify the setback waiting cycles in the run options of the job. When the maximum is reached, the order goes to error state. Sometimes it is desired, that the order goes to suspend instead because you want to have access to the order to relaunch it for a next try.
Solution: This is possible by defining an error node for the node with a dummy job, which exits with "exit 1". This error node is defined as "on error suspend". If you now want to relaunch the order in the error node, you can first set the state of the order to the node with the "on error setback" error handling and then you can resume the order.