If a script has an error, we should rollback the eventually changes made to cassandra to the version before the execution. This is also important because we retry scripts in case Spark has some not managed errors #57. Anyway in case of errors, the script should stop immediately and avoid writing anything on the database (e.g., maybe number_of_process_instances does write anyway).
If a script has an error, we should rollback the eventually changes made to cassandra to the version before the execution. This is also important because we retry scripts in case Spark has some not managed errors #57. Anyway in case of errors, the script should stop immediately and avoid writing anything on the database (e.g., maybe
number_of_process_instancesdoes write anyway).