A system administrator runs a Jython script that creates and configures several servers and realize that the script fails before completing. Although many tasks succeeded, they are rolled back when the script fails.

Before running the script again, what can the administrator add to the script after each task so the completed tasks are committed even if other tasks fail?