Archives of the TeradataForum

Message Posted: Fri, 19 Apr 2002 @ 14:57:41 GMT

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

<-- Anonymously Posted: Friday, April 19, 2002 10:51 -->

I reckon that Paul and Neil are both correct in that NULLs in a primary key will affect RI and joins.

However, using special values can be risky. It is possible that when the data is being queried, that the special value may be interpreted as a real value and therefore the information is distorted.

It was mentioned that the source systems cannot provide all of the data. Could you stage the data into a temporary table and then populate the NULL columns when you load the data in from other sources. When and only when these nulls are replaced with real values, the data can be moved to the intended destination.

I recognise that this may not be possible in your environment but you should tread with caution when using NULLs in a primary index and even more so in a primary key.

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