Archives of the TeradataForum

Message Posted: Tue, 12 Sep 2000 @ 20:43:54 GMT

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

Roland,

It's very probably not the USI itself, but more likely to be the load process or the NUPI.

What utility/program are you using to load the data?

How much data (number of rows and row size)?

How many rows per PI data value?

As Carlos suggested, a fairly common approach when loading is hamepered by SI processing is to drop the SI first, load the data and then re-create the SI, but I'm guessing that may not be an option for you as (presumably) the USi is being used for duplicate checking.

An alternative is to use the load utility to force an error on rows which break your duplicate primary key rule. I've seen this done in Multiload by forcig the value of a BYTEINT column to be outside the acceptable range. That way you can drop the USI, which means that you can use Multiload to load the data, which may be faster for you.

It all really depends on what your requirements are.

Regards,

Dave

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