Home Page for the TeradataForum

Archives of the TeradataForum

Message Posted: Thu, 10 Jun 2004 @ 15:48:41 GMT

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

Subj:   Re: Performance Improvement on this Query without Collecting Statistics
From:   Victor Sokovin

  Is there any other way to improve Performance of the below query other than COLLECTING STATISTICS .....There are around 180 Million rows in DAILY_SPEND and 882 rows in DATE table....  

It might be worth experimenting with PI structures on both tables. There was a discussion recently on AMP-local and cross-AMP aggregation processes which might be relevant for this example. The thread should be in the May archives, if memory serves.

SIs on DateID might be beneficial for that particular query but you should not forget that they will not improve performance of your DML and will require extra space and administration, including collection of statistics etc.

If you are sure that the logic of the application eliminates duplicates under all circumstances you can define BCWK_SPEND as MULTISET. That will improve the insert part of the query.

Some applications might consider using JI with or instead of such queries.

As a side remark I'd mention that using DATE as a table name might create some problems because DATE is a reserved word in TD but this has nothing to do with performance.



  <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: 15 Jun 2023