Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Wed, 14 Aug 2002 @ 14:54:06 GMT


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


Subj:   Performance of a count(*) query?
 
From:   McBride, Michael

I have a query performance question...

Should the following query (in QueryMan):

Select count(*) from aeo_proddb.daily_store_inventory;

take nearely 5 minutes to complete?? (00:04:54 to be exact)

The table currently has 574,802,148 rows

The tablesize is 53,256,011,264 bytes (approx. 50 Gb) and contains daily inventory data for the last 5 and half months...

The table has a non-unique primary index (NUPI) on three columns (date, store, sku)

The table is very evenly distributed across 20 vprocs (zero (0) table skew factor)

The system is a two-node 4850, each node has 4 - 700 Mhz processors and 4 Gb of ram memory (1 Gb per processor)

The database RELEASE is V2R.04.01.00.12

The database VERSION is 04.01.00.58

Out of curiosity, would the count(*) query perform faster if the index was changed to a UPI (date,store,sku are unique across the table and considered to be the primary key in the logical model)

What impact, if any, to ETL time when multiload runs each day (NUPI vs UPI)? Current load times range from 30 minutes to an hour for 4 million rows depending on system load.

Michael E. McBride
Teradata Database Administrator
Teradata Certified Master
American Eagle Outfitters



     
  <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