Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Tue, 21 Nov 2006 @ 11:23:56 GMT


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


Subj:   Re: UTF-8, Fastload and linux
 
From:   David Clough

Just thought I'd close this thread off by reporting back on what we found and what our conclusions on the subject are.

We were somewhat alarmed at the suggestion that Fixed length format would load much, much faster in Fastload than Separated files, so we ran a comparative test :

Input file : 40,000 records (no UTF-8 characters) with a record length of 816 Chars.

Total processor in Fastload (Fixed Width) : 0.4 secs

Total processor time in Fastload (Separated file) : 0.44 secs

This was run on our Test machine, having 9 Amps, with a Fastload Sessions setting of 1.


The 10% overhead in time we feel is quite acceptable.

The file that we were dealing with contains a great number of character based columns with a high degree of column width variability.

Certainly NCR confirmed that we can't use Fixed width definitions for UTF-8 in Fastload, but at least I can also say that it's not possible in Oracle SqlLoader either (thank goodness for that).

We also like the idea of having a physical file on the load Server, rather than use a Unix Pipe. If nothing else, from a call-out perspective, it's much nicer being able to refer to something 'tangible'.

There it is then, we're happy.

Dave Clough
Database Designer
Express ICS

www.tnt.com



     
  <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