How to improve performance with Fast Load on release R19.

Document ID : KB000009682
Last Modified Date : 14/02/2018
Show Technical Document Details
Introduction:

How to improve Fast Load performance when running Load with Output-control Build and Recluster no. 

Background:

Prior to Release R19, Fast Load output-control Build and Recluster No functions differently when there are duplicate or out of sequence keys found, hence the PFL0202E message.  Prior to Release R19, Fast Load would fail when it hits this duplicate or out of sequence key message however, on R19 it compares the keys to eliminate duplicate or out of sequence keys. This process can be time consuming when the table has millions or rows and thousands of out of sequences keys. 

Environment:
Release R19.
Instructions:

In order to improve Fast Load performance, change Recluster No, to Recluster No Sort-Clindx, this will sort the index in clustering order and avoid Fast Load comparing keys when there are duplicate or out of sequence keys. When using Recluster No, your data has to be in clustering order.