Archives of the TeradataForum

Message Posted: Wed, 01 Feb 2006 @ 09:25:56 GMT

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

To the best of my knowledge both should give the same results.

The primary difference between multiset and set tables is in terms of duplicate row checking.

Selects and deletes should work the same way because the above clause is not applicable.

(A)

In a set table with UPI, teradata uses UPI to make sure that there are no duplicates ( if your PI column(s) is unique, you know for sure that your row is not duplicate, so no need to check for entire row, just check the UPI column(s) )

(B)

In a multiset table with UPI, the fact that we have a UPI makes it a necessity for teradata to check for uniqueness of PI column(s)

As you can see the effort/activities invloved in both (A) and (B) are the same so these two tables shouldn't have a difference.

Now that's my theory :)

Joseph Vinish D'silva

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