DSXchange

A new Teradata MLOAD parallel job loads one column of a table from a dataset -- the column is of type/size CHAR(11). The TD MLOAD job reads the file, but wants to truncate the data by one character. Everything across the board is CHAR(11). No truncation is needed. The data gets loaded, but is wrong -- it is missing the final 11th character.

WE get the same results with a sequential file as input.

Here is the warning I'm getting:
TW_PAVE_SKUS,0: Warning: At row 1, link "SEQL_PAVE_PP_INTO_E1DW_WORK_PAVE_SKUS", while processing column "SKU"
Value "1000BBA4106" truncated to "1000BBA410"

Argh!!! Any ideas?

1. Verify that it really is Char(11), not Char(10). (I have to ask.)
2. Verify that there are no non-printing characters in the data.

what is the data format? FASTLOAD or VARTEXT?

This sounds... familiar. Perhaps a carefully chosen search would turn something up.

I had turned unicode on in the column metadata info as I was getting an error about type conversion from string to ustring. When I ran it without the unicode setting on the target columns, it loaded correctly without truncation. Guess there's always more to learn about unicode...

Any thoughts on whether FASTLOAD or VARTEXT would make a difference?