Archives of the TeradataForum
Message Posted: Tue, 29 May 2001 @ 16:17:22 GMT
Assuming yes, I doubt if there's much to choose between the two. Either way, each input record will be sent to the dbms twice (once for each table), so work table sizes will be same.
Only difference I could guess at would be that the option with two APPLY statements may use more host cpu cycles:- APPLY processing is host-side processing and that has traditionally been an area of Multiload which uses lost of cpu cycles. Of course, whether you'll notice a difference depends on the numbers of rows you're processing. You'll probably have to try it.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|