Your job runs slowly and does not scale beyond two nodes. It appears the data has been
partitioned on the DoNotCall flag, which has accumulating totals based on this flag using a sort
method.
Which technique would improve scalability and performance?

A.
Increase the number of physical nodes available for partitioning.

B.
Add an additional column for partitioning to result in additional data partitions.

C.
Change the preserve partitioning option on the stage ahead of the aggregator to clear
partitioning.

D.
Change the partitioning method to Round Robin; add a second Aggregator stage with
Execution mode ‘Sequential’ and

Explanation: