Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Wed, 07 Apr 2004 @ 04:13:56 GMT


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


Subj:   Re: Replace NUSI with JI of same definition?
 
From:   Claybourne Barrineau

Donald,

We've done some pretty extensive work with JIs; however, nothing like you are contemplating. Neat idea!

A few questions:

1) If the optimizer doesn't choose the NUSI, is the LT being redistributed or is a hash join occuring between the LT and spool of the 2 duplicated\product joined STs?

2) How many components make up the UPI of the LT? If only a few, then maybe the use of Key tables hidden in a view will help

3) Is the fact table part of a dependent datamart? If not, the rebuilding (assuming you use multiload to load your LT) of the JI will (potentially) lock all users from accessing the LT for the duration of the build

4) Can you reduce the datablock size for the LT in order to increase the probability that a NUSI will be used (assuming you are OK with slightly longer Full table scans)?

5) Do you plan on Archive\Restoring this LT?


Thanks,

Clay



     
  <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