Archives of the TeradataForum

Message Posted: Tue, 25 Jul 2006 @ 11:50:18 GMT

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

I would recommend that you try creating a single table sparse JI on table 2 with PI COL2

That would remove the impact on table 1 and you would still get reasonable performance on the join

On second thoughts did you try by just having an NUSI on COL1, COL2 for table 2 ?

And yes remember to collect statistics (on columns appearing on where/join conditions).

Joseph Vinish D'silva

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