Archives of the TeradataForum

Message Posted: Tue, 09 Jul 2002 @ 19:31:47 GMT

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

For the past 6 months, we have been loading ~300,000 records per day, with a sequential surrogate key being generated using a combination of a Clock Table (Storing the maximum key value used) and an Insert Select process using the RANK function.

This has proven to be highly efficient except for one point, and that has been the redistribution.

Our data volumes are increasing rapidly as we consolidate all our source systems, and we expect to reach 1 - 1 1/2 million inserts per day.

I expect performance to degrade rapidly as the volumes increase.

I would certainly be interested in discussing peoples experience of generating a FastLoad InMod to do this.

Regards

Matthew Winter

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