Archives of the TeradataForum
Message Posted: Thu, 20 Jul 2000 @ 05:30:47 GMT
The best approach is to make sure you collect all relevant statistics, try to co-location as many tables as possible even if this means choosing a Primary Index very different from the Primary Key even if there are many duplicate PI values (up to 1000 if the row is small, more if you use a Multi-set table)
Try to review your approach to joining commonly joined tables. Small table / Large Table join strategy, merge join strategy, NUSI / USI / PI join strategy.
We would need more info to help you understand your challenges.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|