Archives of the TeradataForum

Message Posted: Mon, 10 Aug 2009 @ 16:52:43 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

For BTEQ and tpump inserts input sorted on the nupi is actually good when you exploit pack and parameter arrays--all rows destined for a particular amp are sent in a single step so would avoid data block I/O--if a request with multiple statements has multiple rows with the same nupi value they would be written out in a single data block I/O on the amp.

	<Prev	Next>		<<First	<Prev	Next>	Last>>