You could have had a more robust script, that only generated an exit status of 0 if absolutely everything was OK. This *is* the answer - you need a better script, one that properly communicates what happened in sqlplus back to DataStage. Most times, when people have this issue it's because they hav...