Throttling and batch limiting for large datasets