Home Page for the TeradataForum

Archives of the TeradataForum

Message Posted: Wed, 17 Jan 2001 @ 16:10:45 GMT

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

Subj:   Re: IDs datatypes for schema design
From:   Jorge LasHeras


According to my own experience, in Teradata it does not matter which is the data type for the attribute(s) used to define the Primary Index (PI), which may be based or not on the attributes of the Primary Key (PK), internal operations are always performed on absolute binary values, but of course these data types must be consistent for same attributes in all tables, otherwise an expensive data conversion for one of the attribute sets must be done prior to join processing, redistributing them on a work table (or spool table as we call it). Character data types are OK for me, no problem, now varchar for a PI sounds as something not usual. Perhaps the recommendation may be due to the star schema or snowflake data model needed/preferred by MicroStrategy, and the lookups. We do support well these data models, but always recommend and prefer normalized ones.

Hope this may be of help.

Jorge (also from NCR).

  <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