How to replicate Unicode data to Teradata using GoldenGate?

Third Party Software

How to replicate Unicode data to Teradata using GoldenGate?

We are trying to replicate data from a set of Oracle databases into a Teradata appliance. Some of the columns in source contain data in Unicode (UTF-8) character set and we are unable to replicate them properly. Following are the platform details -

Source: Oracle 10.2.0.4.0
Target: Teradata 13.00.00.10 (H/W: Teradata 2580)
GoldenGate: 10.4.0.47 Build 001
ODBC Driver for Teradata: 13.00.00.03

Has anyone done this before? Does this require any special settings?
4 REPLIES
Enthusiast

Re: How to replicate Unicode data to Teradata using GoldenGate?

We are tryig to do the same. We get an error when the ODCB.ini characeter set is set to UTF8.
Did you find yoru answer?
Enthusiast

Re: How to replicate Unicode data to Teradata using GoldenGate?

We sucessfully replicated UTF8 characters from Oracle to Teraadata.

On extract server
Oracle version 10.2.0.4.0
GoldenGate version 11.1.1.0.0 Build 078

On Replicat Server
Goldengare Version 10.4.0.19 Build 002
Teradata ODBC 13.00.00.08
Teradata DB 12.00.03.25

Here are the required settings

1) OGG side (source)

On the OGG side please set in the parameter files (for extract, pump and replicat)

setenv (NLS_LANG="AMERICAN_AMERICA.AL32UTF8")

instead of
setenv (NLS_LANG="AMERICAN_AMERICA.UTF8")

2) On the Teradata replication server

2-1) Set the SESSION CHARACTER to UTF16 in the ODBC.
2-2) Add CHARSET UTF-8 to the OGG globals file.
see following note for globals:
How To Create a GLOBALS Parameter on Windows, MVS, or Unix (Doc ID 965278.1)

Enthusiast

Re: How to replicate Unicode data to Teradata using GoldenGate?

also here is how to check the hex values in oracle and Teradata
don't rely on seeing the characters
Oracle (in our case this is UTF8 characterset)
select adr_ln_1_txt, rawtohex( adr_ln_1_txt) from fru_bll
Teradata is UTF16
Select adr_ln_1_txt, CHAR2HEXint(adr_ln_1_txt) from dev_tcy_fraud.FRU_BLL

for example here is a character with the UTF8 and UTF16 values.
Char UTF8 UTf16
日 e697a5 65E5

Re: How to replicate Unicode data to Teradata using GoldenGate?


I have got this error message when replicat data from oracle to teradata.
Error in COLMAP clause. Mismatched types in column mapping (src name [64], targ
name [-9]).
My env is below.
Oracle:
db: 11gR2.
goldengate: Version 11.1.1.0.0 Build 078
table in Oracle:
TEST1 ( ID NUMBER, NAME VARCHAR2(10));

Teradata:
db: 13.00 express
goldengate: Version 11.1.1.0.3 Build 001
Table in Teradata:
test1 (id decimal 4, name varchar 20);