Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Wed, 05 Apr 2006 @ 17:34:25 GMT


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


Subj:   Re: TPUMP Blocking on multiple sessions
 
From:   Victor Sokovin

  I do however wonder if the blocking can still be caused by a relationship of pack size, number of AMPs, and DSW. TPUMP is still writing to a very finite number of AMPs, so if all the rows of the pack happen to the same AMP, the rows and sessions are sequentially blocking each other. In other words is the blocking an I/O issue, and not necessarily a hash collision issue.  


I think that's a good point. There is a per-AMP limit for the number of concurrent hash-level locks. If the limit is exceeded TD may opt to escalate locking to table level in order to prevent "overheating" of AMPs. If I remember correctly, this is a pretty generic scenario, not necessarily specific to TPump, and this is something which we should keep in mind when increasing the number of sessions.

There should be a couple of good threads on the locking level escalation in the archives. If I find them I'll report back.


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