Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Tue, 10 Apr 2012 @ 17:04:06 GMT


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


Subj:   Re: Possible range of data skew
 
From:   Woodrow, Steven

Depends on how you calculate skew factor.

Many people use (CPU consumed by hottest amp) / (Average CPU across all amps).

This would, then, provide a value of 1 for perfectly balanced and a value equal to the number of amps on your system for a perfectly skewed query.

So, on a 50 amp system if the hot amp used 90 seconds and the average across all amps was 90 seconds, the skew factor would be 1.

If it were 90 on the hot amp, and all other amps used 1 CPU second, then the average would be an average is 2.78 and the skew factor would be of 32.

So, the range will be 1 for perfectly balanced and the number of amps on the system would represent the high water mark for skew factor on your system. ?



     
  <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