Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Wed, 25 Jun 2003 @ 11:41:09 GMT


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


Subj:   Re: Partial & Early Group By
 
From:   Claybourne Barrineau

Anomy,

I've reworded and ellaborated on some examples I pulled from a V2R5 presentation...

Early Group By:

TblA (ColA, ColB, Part_No, Vol_Sold)
TblB (ColC, Part_No, Available_Qty)

TblA - 500 million rows and 10,000 unique occurences of Part_No
TblB - 100 million rows and 10,000 unique occurences of Part_No

Select      TblA.Part_No
,     Sum(TblA.Vol_Sold)
,     Sum(TblB.Available_Qty)
From  TblA
,     TblB
Where TblA.Part_No      =     TblB.Part_No
Group By    1

-Pre V2R5, the optimizer would have joined all 500 million rows of the TblA to all 100 million rows of TblB before grouping resulting spool on TblA.Part_No (while summing Vol_Sold and Available_Qty)

-With V2R5, the optimizer will first group TblA on Part_No (while summing Vol_Sold) in parallel (?) with grouping TblB on Part_No (while summing Available_Qty). Once complete, the optimizer will then join the resulting 2 spool files together via Part_No

Partial Group By (I'm not 100% certain about this one):

TblA (ColA, ColB, Part_No, Vol_Sold)
TblC (Part_No, Part_Nm)

TblA - 500 million rows and 10,000 unique occurences of Part_No
TblC - 10,000 rows where Part_No is Primary Key

Select      TblC.Part_No
,     sum(TblA.Vol_Sold)
From  TblA
,     TblC
Where TblA.Part_No      =     TblC.Part_No
GroupB y    1

Pre-V2R5, the optimizer would have joined TblA to TblC on Part_No prior to grouping data by TblC.Part_No and summing TblA.Vol_Sold With V2R5, the optimizer first groups TblA on Part_No while summing Vol_Sold, then joins the resulting spool file to TblC on Part_No


Hope this helps (and is accurate)

Clay



     
  <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