Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Mon, 31 Oct 2011 @ 11:14:32 GMT


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


Subj:   Re: Teradata performance Stabilization
 
From:   McCall, Glenn David

Put simply the TPerf relates to the amount of the hardware you have (e.g. 1 node, 2 nodes etc) and the hardware type (e.g. 5600, 5650) etc

**However**, system 2 is a higher TPerf so is obviously a more powerful system. Perhaps it is running more work than System 1. Have you looked at workload optimisation? Have you engaged Teradata to conduct a review?

Perhaps System 2 has so much work running on it that it is thrashing ( en.wikipedia.org/wiki/Thrashing_(computer_science) )?

At a site I was recently at, some queries that were consuming thousands of hours of CPU hours (TPerf Hours) were improved (with some simple changes) to consume tens of hours (i.e. > 100 x improvement). Similarly through workload management some very simply queries (e.g. Database x;) response times increased by similar amounts. For example the longest my database x; query took to run was 49 1/2 minutes - yes minutes, not seconds, minutes! Now the same query is virtually instantaneous and the system is running more work than ever before.

Some Teradata consultants looked at the workload as a whole and setup workload management. Some others looked at the worst of the worst queries and tuned them and provided advice to the authors of the queries/systems/reports from which the queries originated.


Hope this helps.



     
  <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