Archives of the TeradataForum

Message Posted: Sat, 30 Jun 2007 @ 14:57:59 GMT


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

Subj:		Re: How to get confidence on derived spool

From:		Victor Sokovin

I have a major volume of data in history (DBviews.stmnt_sub) tables with partition on date range.

I have a scenario to take 3 partitions data into 2 different spools. And do the left outer join between these two spools.

I'm getting no confidence on Join spool. Pls provide best practice solution to active this prob.

There is a thread on "no confidence" and joins in the archives:

www.teradataforum.com/teradata/20040121_141429.htm

And I think Clay's formulation still stands: "Also, the left outer join to the derived table will result in no confidence (because the optimizer will have no stats on result of the derived table)."

Of course, we are three years ahead of that thread now and can probably expect some improvement but I can't recall anything particular in this area which is supposed to guarantee more "confidence".

Regards,

Victor