Character Set Conversions (Encoding) Cause Data Loss in ODI Integration Interfaces Loading Microsoft SQLServer Datastores (Doc ID 966637.1)

Last updated on OCTOBER 02, 2023

Applies to:

Symptoms

When attempting to load data into a Microsoft SQLServer 2005 Datastore using an Oracle Data Integrator (ODI) flow:

• From an Oracle table present in a database schema with the UTF8 character set (NLS_CHARACTERSET)
            
• To a Microsoft SQLServer database table created with the SQL_Latin1_General_CP1_CI_AS collation (Microsoft term to designate the character set).

When checking the data in the target table, it appears that some characters have been lost and are represented as question marks (?).

For example, multi-byte data 원 용 선 present in the Oracle table is translated to ? ? ? after being loaded to Microsoft SQLServer.

Why does this happen?

Note that in the Microsoft SQLServer table, CHAR and VARCHAR data types have been used.

Cause