Hi,
I think the major issue when I import data from Excel, Text, or Access
into the database on SQL Server, if the error occurs, I really cannot tell
what record(s) is causing the error of the package. Is there anyway I can
find out the bad record(s) casuing the error???
I used to check on the "Always commit the final batch" and set the "Commit
Batch Size" to 1 so that I can tell what would be the last record to commit
successfully and look at the next record in the source file to find the
caused error one. Is there any better way of doing this?

Thanks
Ed

If you are using SQL Server 2000 have you looked at the last tab of the
datapump task for logging of errors? It will allow you to log failing rows.
I have not always managed to get this to work though so your mileage may
vary.

Another way and it is a long way as well would be to note the error because
that should tell you the attribute name and the type of error. You can then
write queries against your source to detect the rows. Text files will need
to be imported to a scratch working table first of course.



"Ed" <Ed (AT) discussions (DOT) microsoft.com> wrote

The issue for the first method is it is not always tell me the correct source
row number that causes the error...
and if there are more than one errors happen in the source file, i don't see
the error log file can log more than one rows....
Ed

"Allan Mitchell" wrote: