Archives of the TeradataForum

Message Posted: Tue, 28 Jan 2014 @ 13:54:02 GMT

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

The SQL logic is: small_table left join ( big_table inner join very_big_table)

Which plan do you find better ?

1. ( small_table left join big_table) inner join very_big_table ?

Result will be different - very_big_table will potentially eliminate some rows from small_table

2. ( small_table left join big table) left join very_big_table ?

Result will be different - very_big_table will not eliminate rows from big_table.

You know, that big_table is subset of very_big_table, but Teradata doesn't.

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