Archives of the TeradataForum

Message Posted: Tue, 14 Mar 2006 @ 11:05:28 GMT

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

You can do it, but I wouldn't recommend it. Access will tend to run a Teradata query for each row of the local Access table. The Access optimizer isn't client-server aware, and it optimizes around Access.

I've used a couple of different approaches, depending on how much data is on the Access vs Teradata side, and how many queries need to run. If the Access db needs a reasonably sized subset of the data, create a passthrough query to select the Teradata rows, then use the passthrough query as a source for an Access append query. You may need to write a module to generate the appropriate SQL. Once the data is local to Access they can massage it at will.

If the Access data is reasonably small then create a working table in Teradata and upload the Access rows. This approach is useful when there's some user defined filter or segmentation data and most of the queries run on the Teradata side. Don't forget to define a reasonable PI on Teradata, otherwise it defaults to the first column.

If the Access data set is large you may need to export it to text and use multiload or fastload to load Teradata.

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