Archives of the TeradataForum

Message Posted: Thu, 06 Nov 2003 @ 10:54:36 GMT

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

There are few things you should consider.

- If you are planning to convert the the column to unicode make sure that you change the character set for this column that reside in any table.

- If this column is joined with a column of other character set (like latin etc) then it will lead to full table scan due to the character translation. Therefore, you might need to change the character set of all columns that are being joined with this column.

- The column will store the data as two-byte character, that is, one character will consume 2 bytes of disk space on Teradata.

- Teradata stores data in ASCII. Before transferring data to client it is translated into unicode or the to the character set that is defined using any extraction utility.

- The translation takes place at Teradata, and doesn't has much overhead on the performance, As the data is passed through another layer of translation, instead of passing directly to the client.

- ODBC is just a bridge to pass data between Teradata and client. You can specify the character set type in the ODBC configuration options for Teradata.

Regards,

Muhammad Altaf
Teradata Certified Master
Teradata, A division of NCR

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