Archives of the TeradataForum

Message Posted: Mon, 14 Feb 2005 @ 19:02:00 GMT

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

This topic is showing up on the forum quite often. Some posters report success with such loads and even post their scripts but others seem to have problems making the scripts work on their environments.

My understanding is that the layout of all the records must be similar: same number of fields in each record. It might be easier to achieve this with the CSV format.

If that is true, input files need to be manipulated before their submission to ML. Here my interest in the topic usually ends because if I have to manipulate the files I may just as well split them as I wish and load separate parts using simple ML scripts.

Error checking might in principle get more complex with this approach. If so, the information in headers can be just added to the main records. A bit redundant perhaps but simple to handle as you will need only one layout in the ML script.

Regards,

Victor

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