Alright well, thanks for the input. I think we may play with the scheduling of the job/tuning the job to see if we can free up enough resources to avoid...