Archives of the TeradataForum

Message Posted: Wed, 26 Feb 2014 @ 10:33:08 GMT

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

Thanks Dave Wellman and Barry Hull for the suggestions - both do indeed work.

What I'd missed on the SET Table approach was that, when defined as 'NO LOG', the second Insert removes the first Inserted row - weird but true. Surely enough, when defined as a Logging Table, it ignores the subsequent duplicate rows.

As it happens, the Atomic Upsert suits my purpose very well, so I'll go for that option.

Thanks again for the excellent suggestions.

Regards

David Clough
Senior BI Database Designer
BI Competency Centre

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