Archives of the TeradataForum

Message Posted: Mon, 31 May 2010 @ 13:37:00 GMT

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

Not sure if this is more of an Oracle question than a Teradata one, so you might just want to delete this note out of pure disgust ! (I'd quite understand).

Anyway, in migrating data from our operational system (Oracle) to our EDW (Teradata) we have found the odd (and I do mean 'odd') characters that, whilst Oracle accepts them, Teradata fails on Load (Fastload).

I've got the Teradata side defined as Unicode (at Column level, obviously) and the Oracle side is defined as Unicode (at the database side - yes, I know, rubbish!).

It seems to me as as Oracle is being more forgiving in which characters it will accept, whilst Teradata (correctly in my view) rejects them if they're not in the Unicode character set.

Has anyone got any experience of this type of situation and how to handle it?

I'm already handling extraneous Carriage Control Line Feed issues that creep into Oracle by calling a function to strip them off before unloading the data (i.e .before it hits the file system). The reason for this is that we're using Fastload with variable length delimited files.

I can get examples of these characters if anyone's got any ideas of how to resolve the situation.

Thing is, the extract from Oracle works and then the Teradata side fails completely, all for the sake of one pesky little character !

Regards

David Clough
Database Developer
Database Design Group

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