I noticed contents in TEXT/NTEXT columns from error rows being added to the data in columns in the next good record. Best example of this is to have a flat file with columns declared as COL 0, int, COL 1 text and COL 2, varchar(10). The
1, Test this one, Passed
A, Test this two, Passed
B, Test this three, Passed
2, Test this four, Passed
3, Test this five, Passed
The two rows with A and B in COL 0 will cause an invalid datatype error and be sent to the error table/file via redirection. When the next row with 2 in COL 0 goes to the destination, the COL 1 text field will be "Test this two Test this three Test
this four" instead if "Test this four". This is a simple example and everyone said use varchar/nvarchar, but I have strings that are over 30,000 in length. What is an alternative to this problem?
Microsoft is conducting an online survey to understand your opinion of the Technet Web site. If you choose to participate, the online survey will be presented to you when you leave the Technet Web site.