I have been having difficulties for many months now when importing text data
that is exported from SAP then FTP's from a UNIX server to my SQL server.

What is happening is, when I use the text file source and data pump to
import the data, I do not receive all the rows.

I can take the same text files and import them into either Access or XL
without any problems. This is actually how I found out that the rows were
missing.

I have looked at the text files with a HEX editor and the row indicator is
0D and 0A (carraige return and line feed).

I also noticed that If I am importing a fixed field file and the file
contains 6 columns and the first row contains data only in the first 5
columns, you can not parse or define the 6th column. To remedy this, I
copied a record in the file that contained data in all 6 column and moved it
to the first row. When I did this, I was able to define all 6 columns.

This problem has haunted me for many months now but only when the data comes
from SAP. I work for a Koren company where almost every machine contains
multilanguage support for Koren. Im not sure it this is the problem or if I
am grasping at straws here, but it is the only thing I can think of.

I do not want to load multi language support on may local maching or server.

Many of my data feeds have to be manually imported into Access first then
into SQL.

Please.... Please..... help me solve this problem.

Gary,

Can the issue be related to ftp transfer mode? Have you tried binary as it
would not change row delimiters around?

Ilya

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

Yes, I do have it set to binary.

I'm not sure if it is a bug or not, but I feel it is related to the data in
the last column. As mentioned below, If the structure is initally defined as
6 columns, but the 6th column of the first record does not contain data, it
appears that it not bringing in remaing records that do have data in the 6th
column.



"Ilya Margolin" wrote: