Archives of the TeradataForum

Message Posted: Thu, 26 Sep 2003 @ 03:20:38 GMT

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

I am not really sure why this should be bad. As far as I recall there is nothing in relational DB theory that fundamentally prohibits the update of PKs. I understand this may be expensive in terms of effort (e.g., cascading ripple effects in some DB products, redistribution in others...)

Also, I am not really sure why the approach of deleting and inserting should be so much better. If it is, I would strongly expect NCR to consider a similar approach. After all it is better to have the system be smart in standard cases.

Am I not seeing the tree for the forest?

Thomas

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