Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Fri, 13 Jun 2003 @ 14:28:10 GMT


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


Subj:   Re: PPI's & hard-coded predicates
 
From:   Charles, Alan

Hi Yonina,

The same, i'm afraid.

The problem seems to be that unless the optimiser can see exactly what it is dealing with when it builds the explain (i.e. unless the predicate is hard-coded), it just does a FTS. This is not really a total surprise since a join (or subquery) could potentially return any number of values, and none of these actual values are available to the optimiser at the time of determining the best plan. If there's only one row on the control table, the column's a UPI and statistics are collected then I suppose it might be reasonable to expect the optimiser to detect that there'll only be one partition to deal with. It doesn't though, unfortunately!

I actually don't think there is any other solution other than variations of the two options I originally stated. However, I just thought I ought to clarify my understanding before I go public with it!


Regards,

Alan.



     
  <Prev Next>   <<First <Prev Next> Last>>  
 
 
 
 
 
 
 
 
  
  Top Home Privacy Feedback  
 
 
Copyright for the TeradataForum (TDATA-L), Manta BlueSky    
Copyright 2016 - All Rights Reserved    
Last Modified: 27 Dec 2016