Teradata 6706 error

shravya346 · Post by **shravya346** » Wed Dec 07, 2011 3:08 pm

when i am trying to load a mainframe file to teradata , i am getting the following error :

Teradata_Connector_6,0: RDBMS code 6706: Field CARD_TRACK1 (CC_TeraAdapter::selectLoadErrorTable1, file CC_TeraAdapter.cpp, line 7,430)

I am able to load the same file into oracle table but not in TD .

the error indicates untranslatable character in the data .

What do i need to do for the data to be loaded into the TD table .....I have tried changing the field to UNICODE character in TD table definition and even then its failing.

ray.wurlod · Post by **ray.wurlod** » Wed Dec 07, 2011 3:12 pm

Welcome aboard.

What method (that is, what stage type and what settings) are you using to write to Teradata?

shravya346 · Post by **shravya346** » Wed Dec 07, 2011 3:18 pm

I am using

seq file >>copy>>TD connector

TD Connector :
Access Method : BULK

shravya346 · Post by **shravya346** » Thu Dec 08, 2011 6:14 pm

How do i trim these special characters in datastage , is there any way of doing this in DS , if not how can i handle that while extracting from oracle database .

Aruna Gutti · Post by **Aruna Gutti** » Fri Dec 16, 2011 11:28 am

Hi Sravya,

What version or Teradata you are using?

We set up our teradata connector stage with following options and the special characters you showed in your post get loaded through DataStage.

Variant:12
Transaction Mode: ANSI
Client Character Set: LATIN1_0A
Automap Characterset coding: Yes

We configured our database to accept LATIN1_0A character set.

Hope this helps.

Aruna.

kwwilliams · Post by **kwwilliams** » Fri Dec 16, 2011 1:31 pm

Your Parallel NLS Map should be set to ISO-8859-1, which will map cleanly to LATIN1_01A. And all of your character columns should NOT have extended properties set to Unicode.

If your source is Unicode, you need to perform a Ustring to String conversion, before landing to teradata. That should clear up your bad character UNLESS the bad character exists in source data.