Archives of the TeradataForum

Message Posted: Fri, 30 Nov 2007 @ 14:34:46 GMT

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

One of the reasons is that running two or more synchronized (not replicated bit by bit) live systems may become fashionable in the (near) future. To be able to have the same IDs you must generate them using some mathematical functions on the business keys. So, same business keys = same surrogate keys on all databases.

My personal opinion: I don't like surrogate keys. They disturb my user experience and although they are tempting from the design point of view I'd rather not use them.

But sometimes there are cases when you don't have a choice. You receive some unstructured data feed from a third party, for example, and there is no natural business key to use.

Regards,

Victor

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