Sybase IQ Load Table Error

[Ashish B]

Hi Friends,

Technology : Sybase IQ 12.6
Enviornment : Win2003

I am facing issues while loading table. I am geeting a error message as
below -

ASA Error : 1013027
Sybase Error Code = 21
SQLStatus = "QCA27"
Error Message - "Number of bytes (512) for a column from an input file has
exceeded the maximum allowed (124)"

Source and target table has 103 column all with the Varchar (255).

Any help on this is appreciated.

Thanks in advance.

Regards,
Ashish

[Leonid Gvirtz]

Hi

In my experience, this kind of errors is usually caused by using invalid
input files (malformed rows with more column delimiters than needed, for
example) or by using wrong column/row delimiters in the LOAD TABLE
statement. Did this kind of load work previously or it is a new one that
never worked before?

In such cases I usually start from verifying that column delimiters and
row delimiters are indeed what I expect them to be via a hex editor.
This approach is especially useful if you use special characters (\n, \r
etc.) as delimiters.

Hope it helps
Leonid Gvirtz

[Ashish B]

Hi Leonid,

Thanks for the reply. I checked the row and column delimeter, they are
exact. I tried to limit nos. of rows - only 5 olumns with only 10 rows for
the said table.It woked fine could see data in file.

Same structured ( with same row and column delimiter ) flat files has been
loaded successfully.

Is there any other reason behind this.

Regards,
Ashish

"Leonid Gvirtz" <lgv...@yahoo.com> wrote in message
news:49c34692@forums-1-dub...

[Leonid Gvirtz]

Hi

Did you try to load only a portion of the problematic file (all columns
included)? If it works then the problem may be related to the data in a
number of rows in the problematic input file and not to the number
columns or rows that you load. You can split your file into smaller
portions (by script) and narrow down the problem this way.

Some possible scenarios that produced similar errors in my experience:

-- You may occasionally have column/row delimiters inside your data.
-- The data contains NULLs of some kind which are incorrectly
interpreted by the LOAD
-- The data contains datetime of incorrect or unexpected format. But it
is not your case if all your columns are varchar(255)

If you have IQ 12.7 installed on your site that you may try to load your
file into a test table via iq_bcp, it may help to find the problematic
data. Alternatively, if you have ASE on your site that you may try to
"bcp in" the data into ASE and see what happens. Particularly, look for
rows that are apparently incorrectly loaded.

As far as I understand, the load of the problematic file doesn't even
start. But, if it does start, you can specify "NOTIFY 1" or MESSAGE
LOG/ROW LOG clauses in the LOAD TABLE statement, then finding of
problematic rows should be easy.

Hope it helps
Leonid Gvirtz

[ignaciovera]

Hi,

My experience with the load command is somewhat confusing.
We are producing some load files inside a python script and
loading into IQ. So far so good and everything works as
expected.

If I try to load the files directly using a client (I have
tried three different clients) and using the same command
used inside the python script I got the same error you are
mentioning in this thread.

I haven't investigated further but my feeling is that there
is a conflict with the encoding. We are using special
characters (\n, \r) as delimiters.

Cheers,

Ignacio

[Ted]

if the file is not too large you can always run an octal dump and
check for control characters ( od -c x | grep \\n)

I have run into the same problem many times and the issue has always
been a special character.

[Shamim Aziz]

Hi,
Is this Load table is written in procedure or directly executing in SQL
Advantage.

Regards
Shamim Aziz
<Ignacio Vera> wrote in message news:49c7638c.1ad...@sybase.com...