Archives of the TeradataForum

Message Posted: Wed, 28 Aug 2002 @ 09:26:01 GMT

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

Don't forget when choosing a compression value for a column that NULL is automatically compressed in addition to the chosen constant value. This means that even though the most frequently occurring value may be present on a modest % of rows (say <50%), when added to the frequency of NULL values for the column it may make a suitable compression candidate. This was briefly mentioned in Dave Wellman's reply.

I typically end up running a lot of "SELECT COLUMN_NAME,COUNT(*) FROM TAB_X GROUP BY 1 ORDER BY 2 DESC;" type queries against the obvious candidates in the largest tables for this one.

For those on later releases scan times, but not table sizes, can also be reduced by using the 128Kb block sizes available for tables.

Roll on V2R5!

Paul Johnson
Certified Teradata Master
Leading Edge Business Solutions Ltd

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