However even if it was 50k or 100k records (Which is totally possible from only 1 of our vendors) I need to understand the execution process of a MR script and if it will restart at the start of the dataset if the script is halted because of limits