Archives of the TeradataForum
Message Posted: Tue, 13 Aug 2002 @ 19:53:21 GMT
Regarding the use of 'CSUM(1,1)' to generate a unique row ID: all data will be loaded to a single AMP for processing; therefore, don't be surprised if you blow out of spool space if the table you are working with is very large. If a Primary Key exists, then use 'CSUM(1,Primary Key)' to generate a unique row ID. It is much faster because of the even data distribution.
My 2 Cents,
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|