Archives of the TeradataForum

Message Posted: Thu, 07 Oct 2010 @ 15:18:50 GMT

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

The error is not telling you that the row is too large it is saying that the DDL text itself is too large. (approx 113K in this case)

The suggestion to change from VARCHAR to CHAR would reduce the size a little (but not enough) and greatly reduce both the size of the data rows and their performance.

If you define your default character set as UNICODE you can remove the 'CHARACTER SET UNICODE' phrase from every column definition. That would bring the DDL size down below 64k.

You can also use CS instead of CASESPECIFIC to reduce the size further. (Depending on whether you use ANSI or Teradata mode your default may be NOT CASESPECIFIC so you could remove that phrase entirely.)

Mike Dempsey
Teradata Client Tools

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