Teradata Mload Vs FastLoad Vs TPump


Teradata Mload Vs FastLoad Vs TPump

Hi All,

I am new to Teradata and to this forum. Some questions popped up in my mind when I was understanding teradata architecture and data loading tools.

  1. The difference between Mload & Fastload is like, Mload loads data first to Work table and then to the actual table, Where as Fastload loads first moves data to all AMP and then the hashing takes place. But , it was also mentioned that after the hashing is done , row hash will tell the AMP number in hash map. I am finding these two statements contradicting. Where will hashing take place actually? and how it works exactly in mload and fastload?
  2. If all the rows are moved to all AMPs first in Fastload, then after hashing lofgically only a few rows will fit into a particular AMP. What will happen to the remaining rows that are moved to that AMP? Is it like those wil be discarded as same rows are present on each AMP or will those be moved to the corresponding AMP (i.e. AMP to AMP communication)

Can some one plz help me in understanding these concepts ?


Sai Pavan Kumar Bhamidipati