We are trying to load a fixed length file into a DB2 table. We have set all the fields in the table to be of the type VARCHAR, and we know the lengths are greater than the source data.

No matter what we try, we keep getting SQL0433N errors - Value " " is too long.

It seems to me that DB2 UDB EE is potentially determining the datatype of a given set of data (say positions 1-10) by looking at the data contained, and setting up a page with datatype set. While this might explain data whose first set of values is numeric in positions 1-10, followed by alphas in the same positions, this does not change the idea that total length of a piece of data is set by the position boundary (1-10, 11-20, etc....)

Am I missing something here?