DSXchange

Hi All,

What is the most optimum way of splitting one input into 3 output files based on the filed in the input.

Input is being read from the Teradata table and the output should go into 3 different files based on the value in one column of the input data that is read from Teradata.

Thanks in advance

Kishore

If there are only three (or few) distinct values in the column, Switch stage.

If you're reading in sequential mode from Teradata consider using a three-node configuration file and a modulus partitioning algortihm. You may need to map the three values to integers using a Modify stage. Hash partitioning can not be guaranteed to work. Write to a Sequential File stage that writes to precisely three files - one should be written from each partition.

Otherwise investigate Filter or Transformer stage, each of which has a computation overhead.