Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Thu, 30 Sep 2004 @ 12:00:34 GMT


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


Subj:   Re: Extra secondary index on large table
 
From:   Victor Sokovin

  We know there will be an overhead to adding another secondary index, but is it possible to estimate what this might be? Obviously, our ideal would be to copy the table, or even half of it, add the index and test both user queries and loads (we are worried our loads may slow down considerably), however, we do not have sufficient space on our system to do this at present.  



You are right that loads might sometimes go wild after a new SI has been added. DML might then take *weeks* to complete. Nobody can wait that long so be prepared to kill the session as soon as you realize that things go wrong. TD will launch the rollback. It can take 1.5-2 times longer than the session itself. Under unusual circumstances the rollback might take even longer. It makes sense to make yourself familiar with rollback killing technique if you are on R5 (on lower releases, you'll need NCR support). It is well documented in the DBA documentation and there are examples in the archives of this forum.

Killing the rollback might corrupt the data in the table so, in the worst case, you have to allow for the restoration of the data. I think if you use JI you are well prepared for such restorations anyway. Trying all this for the first time in production on a weekend might make things less exciting. If you killed the session fast enough you might just want to wait until the rollback normally finishes.


Regards,

Victor



     
  <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