ORA-24347: Warning of a NULL column in an aggregate function

The following query works fine, normally.
select max(ship_date) Ship_date,msn.attribute7 from
     mtl_serial_numbers msn,
     mtl_item_categories mic,
     mtl_categories mc
               where      msn.inventory_item_id = mic.inventory_item_id and
               mic.category_id=mc.category_id and
               mic.category_set_id=1 and
          mc.segment1 in ('ILMI','ILM','ILMI-UNT')
          Group by msn.attribute7
But when I include the same query inside a SQL report it gives me the Following error message
ORA-24347: Warning of a NULL column in an aggregate function
Help me out.

I tried out with some Dummy Date, it worked fine.
But the problem I am facing here is I cannot assume
any Date for the corresponding column. Whether there
is any other possibility to avoid this problem without
giving any dummy date?
thanks,
chandru

Similar Messages

ORA-24347 with Powerbuilder

Found that the column alias is what causes this error when accessing views.
Views that have been created using this method:
select sum(col_11) total, acct_num from abc
or
select sum(col_11) as total, acct_num from abc
will not work with Powerbuilder 8 at present.
It appears that the column alias is handled somewhat different in 8i and Powerbuilder has not supported all methods.

I have also run into a problem with this error message.
In our case although it came to light when we upgraded to PB 8.
Accessing a view triggered the error.
I have found that trying to display the properties of the view in the database painter will bring up a blank messagebox and the view definition will display nothing.
Only if you retrieve data from the same view it will cause a corruption in the connection and any other view or table will display this message ORA-24347 "Warning of a null column in an aggregate function"
In hitting views at random through the database painter - several views were found to suffer from this problem.
There is nothing within the SQL that can explain it.
One of the views is no more than a straight select from a single table.

ORA-01400 - not null column with default value and item with authorization

I've searched - I would think someone has run into this. APEX 3.0.0.00.20 - I've created a simple form on a table. One of the column is a not null column with a default value. I have a select list on that item, but it has security on it - authorization scheme. So, it checks the user and if that user isn't of the right role, it will not even display that item. However, APEX appears to still send in the column in its sql! So, the default value is useless, it sends in null each time. Even if I set the default at the Item level, I get null. Argg. That's got to be a bug...
In debug, I do not see the item listed at all. It's not used. That's fine - but why is it trying to insert the value? I would think it would leave it off??? I think because the item is associated with a database column. But, getting around this is ugly. Having to create a hidden item for each one, and then check to see if I need to take the list value... horrible. Any way to get around this???

I should add - I guess I can always put my own custom process in to replace the DML. Just seems like a simple thing - if the value doesn't appear on the debug, isn't set with any default value... don't include it in the DML.

Oracle BPEL - Does not cater for not null columns and use of "default".

Oracle BPEL - Does not cater for not null columns and use of "default".
BPEL fails with message:
ORA-01400: cannot insert NULL into ("EDDB"."SEISMIC_LINES"."COORD_SYSTEM_ID")
But SQL*PLUS command works:
INSERT into EDDB.SEISMIC_LINES
(etc)
regards
Allan Ford
Analyst / Programmer - IT Application Services, IT Services, Shared Business Services
Santos Ltd
Level 4, 91 King William Street, Adelaide SA 5000
Phone: 08 8224 7944 Fax: 08 8218 5320
Email: [email protected]

note: BPEL keeps it's own "offline" copy of table and database items. A column that is marked not null in the database can be marked as nullable in this area. (if you kmow that a trigger is going to cater for this ..)
One workaround is to use a trigger to provide value rather than use the column default ..

I am getting error "ORA-12899: value too large for column".

I am getting error "ORA-12899: value too large for column" after upgrading to 10.2.0.4.0
Field is updating only through trigger with hard coded value.
This happens randomly not everytime.
select * from v$version
Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bi
PL/SQL Release 10.2.0.4.0 - Production
CORE     10.2.0.4.0     Production
TNS for Linux: Version 10.2.0.4.0 - Production
NLSRTL Version 10.2.0.4.0 - Production
Table Structure
desc customer
Name Null? Type
CTRY_CODE NOT NULL CHAR(3 Byte)
CO_CODE NOT NULL CHAR(3 Byte)
CUST_NBR NOT NULL NUMBER(10)
CUST_NAME CHAR(40 Byte)
RECORD_STATUS CHAR(1 Byte)
Trigger on the table
CREATE OR REPLACE TRIGGER CUST_INSUPD
BEFORE INSERT OR UPDATE
ON CUSTOMER FOR EACH ROW
BEGIN
IF INSERTING THEN
:NEW.RECORD_STATUS := 'I';
ELSIF UPDATING THEN
:NEW.RECORD_STATUS := 'U';
END IF;
END;
ERROR at line 1:
ORA-01001: invalid cursor
ORA-06512: at "UPDATE_CUSTOMER", line 1320
ORA-12899: value too large for column "CUSTOMER"."RECORD_STATUS" (actual: 3,
maximum: 1)
ORA-06512: at line 1
Edited by: user4211491 on Nov 25, 2009 9:30 PM
Edited by: user4211491 on Nov 25, 2009 9:32 PM

SQL> create table customer(
2 CTRY_CODE CHAR(3 Byte) not null,
3 CO_CODE CHAR(3 Byte) not null,
4 CUST_NBR NUMBER(10) not null,
5 CUST_NAME CHAR(40 Byte) ,
6 RECORD_STATUS CHAR(1 Byte)
7 );
Table created.
SQL> CREATE OR REPLACE TRIGGER CUST_INSUPD
2 BEFORE INSERT OR UPDATE
3 ON CUSTOMER FOR EACH ROW
4 BEGIN
5 IF INSERTING THEN
6 :NEW.RECORD_STATUS := 'I';
7 ELSIF UPDATING THEN
8 :NEW.RECORD_STATUS := 'U';
9 END IF;
10 END;
11 /
Trigger created.
SQL> insert into customer(CTRY_CODE,CO_CODE,CUST_NBR,CUST_NAME,RECORD_STATUS)
2                values('12','13','1','Mahesh Kaila','UPD');
              values('12','13','1','Mahesh Kaila','UPD')
ERROR at line 2:
ORA-12899: value too large for column "HPVPPM"."CUSTOMER"."RECORD_STATUS"
(actual: 3, maximum: 1)
SQL> insert into customer(CTRY_CODE,CO_CODE,CUST_NBR,CUST_NAME)
2                values('12','13','1','Mahesh Kaila');
1 row created.
SQL> set linesize 200
SQL> select * from customer;
CTR CO_   CUST_NBR CUST_NAME                                R
12 13           1 Mahesh Kaila                             I
SQL> update customer set cust_name='tst';
1 row updated.
SQL> select * from customer;
CTR CO_   CUST_NBR CUST_NAME                                R
12 13           1 tst                                      Urecheck your code once again..somewhere you are using record_status column for insertion or updation.
Ravi Kumar

Warning on a empty column

Hi
I have following requirement:
I have column called Promotional Limit on custom OA Page which is part of advanced table. User can enter the data in this form. I need show a warning if user does not enter value to this column. I'm not sure how I can achiev this.
I tried following:
i) I registered an event with this column, but it looks like this event is not getting fired when user is not entering any data (I realized that fireAction is getting fired whenever(/only when) the field value changes). Since user did not entered any data it did not fire. So, I could not use the code in Controller class.
ii) I tried to write a check in validateEntity() as below to throw a warning:
if (getPromotionalLimit()==null)
throw new OAException("LDMF","SP_PROMO_LIMITS_NOT_ENTERED", null, OAException.WARNING, null);
the above is showing the warning on the page but it should let the user to continue saving the data. It should not stop the user saving the data. Since I'm throwing this in validateEntity() method it is always sending the user back to the page.
So, I need warn the user that he did not enter value to the column, still he should be able to save the data if feels that is OK.
Any ideas?
Thanks
Prasad

Hi,
Or you can try making a Validation View Object (VVO) for the VO that populates the advancedTable, and check if the mapped value is empty there...
Hope it Helps!
Thiago

Interactive Report with Null Columns

I've got a user who wants to export an interactive report, and he needs certain columns to appear on the export for him to fill the data in later (this is an intermediate step, while we work on getting all the data he needs in to the database). I've created the report query, and there are a handful of null/empty columns (to preserve the correct order for the export). When I try to add this query to APEX in an interactive report, I get a "ORA-00001: unique constraint (APEX_040000.WWV_FLOW_WORKSHEET_COLUMNS_UK2) violated" error.
Googling around shows me that this happens because of similarly-named columns, and the solution is to provide aliases to all columns. My report has aliases on all columns, but I still cannot create an interactive report. I've tried changing the null columns to empty strings, as well as enclosing the aliases in double-quotes, but nothing works. I can however use my original query to create a standard report, but due to the export requirement, this isn't ideal.
I was able to create the interactive report with one null column, and then edited the report source to add in the others. This had to be done one at a time, since trying to add multiple null columns at the same time gives the same error again. Unfortunately, when I try and run the page, I get an "ORA-20001: get_dbms_sql_cursor error ORA-00918: column ambiguously defined" error.
My original query:
select customer.customer_name as customer,
       project.name as project_name,
       trunc(project.estimated_end_dt) as due_date,
       project_status.project_status_desc as status,
       null as revenue,
       project.baseline_effort as baseline_hours,
       null as projected_cost,
       null as est_gain_loss,
       project.actual_hours as actual_hours,
       project.estimated_hours as projected_hours,
       null as projected_cost,
       null as projected_gain_loss,
       null as roi
    from project,
         customer
where customer.customer_id = project.customer_id
and project.inactive_ind = 0
and project.customer_id is not null
and project.estimated_end_dt >= :DTP_P50_STARTDT
and project.estimated_end_dt <= :DTP_P50_ENDDT
order by customer.customer_name,
         project.estimated_end_dt;             Can someone tell me a way to create an interactive report with multiple null columns?

Hi shimmoril,
The problem is likely that you have two columns aliased as "projected_cost" (7th column and 11th column).
Hope this helps,
John
If you find this information useful, please mark the post "helpful" or "correct" so that others may benefit as well.*

Rawtohex - How to Insert ? ORA-12899: value too large for column

Hi,
Can any one please help me to resolve the following issue ?
Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bi
PL/SQL Release 10.2.0.4.0 - Production
Name Null? Type
ABC_OID NOT NULL RAW(8)
ABC_NAME NOT NULL VARCHAR2(30 CHAR)
UPDATE_TIME NOT NULL DATE
UPDATE_BY_WORKER_NO NOT NULL NUMBER
I'm able to insert 1st 2 records but when am inseerting 3rd one am getting error:-
insert into caps.ABC_LOOKUP values( rawtohex('SERIES'), 'SERIES','19-FEB-09','1065449')
insert into caps.ABC_LOOKUP values(rawtohex('FAMILY'),'FAMILY','19-FEB-09','1065449')
Insert into caps.ABC_LOOKUP values(rawtohex('CONNECTOR'),'CONNECTOR','19-FEB-09','1065449')
ERROR at line 1:
ORA-12899: value too large for column
"XYZ"."ABC_LOOKUP"."ABC_OID" (actual: 9, maximum: 8)
Thanks in Advance.....

Yes, Done...
Actually I suggested same thing to them (application), But they did not agree with me then i got confused :-)
Now the same thing worked well..Thanks a lot for your time

Rs.updateBoolean SQLException: ORA-12899: value too large for column

Complete error is SQLException: ORA-12899: value too large for column "SMSUSER"."PRUEBA"."VLOGIC" (actual: 4, maximum: 1)
Let's see the code:
PreparedStatement ps=null;
ResultSet rs=null;
try
ps=conn.prepareStatement("create table prueba(name varchar2(32),vlogic char(1) not null check(vlogic in (0,1)))");
ps.execute();
logger.info("Table created.");
ps=conn.prepareStatement("insert into prueba (name,vlogic) values ('user01',?)");
ps.setBoolean(1,true);
ps.executeUpdate();
logger.info("Data Inserted.");
ps=conn.prepareStatement("update prueba set vlogic=? where name=?");
ps.setBoolean(1,false);
ps.setString(2,"user01");
ps.executeUpdate();
logger.info("Data Updated.");
; Till here all runs ok, but if we try to modify vía Resulset.....
ps=conn.prepareStatement("select vlogic from prueba where name=? for update", ResultSet.TYPE_SCROLL_SENSITIVE, ResultSet.CONCUR_UPDATABLE, ResultSet.CLOSE_CURSORS_AT_COMMIT);
ps.setString(1,"user01");
rs=ps.executeQuery();
if (rs.next())
logger.info("Got record.");
rs.updateBoolean("vlogic",true);
rs.updateRow();
logger.info("Column updated.");
catch (SQLException E)
logger.info("SQLException: "+E.getMessage());
finally
closeResultSet(rs);
closePreparedStatement(ps);
The trouble is that when updating via resultset, what is going to be inserted is "true" or "false" and not "0" or "1" as with insertions and modifications via preparedStatements.
So systems returns error: SQLException: ORA-12899: value too large for column "SMSUSER"."PRUEBA"."VLOGIC" (actual: 4, maximum: 1)
Cause it is tryng to insert "true".
Can somebody tell me what's happenign here?
Thanks in advance.
Francisco Javier Ascanio Suárez.
E-mail: [email protected]

Ok, but why is this behaviour different in ResultSet statements than in Prepared Statements?
As you can see in my example, prepared statements with set boolean runs ok.
I like your "proper way", and it resolves my trouble, but it don't tells me why do I have to program a field update in different ways depending of Prepared Statements or updating resultsets.
Thanks in advance.

XMLGEN: Produce XML dump of a table WITH tags for null column values

I am new to generating XML so bear with me....
We have a customer who needs an XML extract of data, including tags for any column that is null.
I created a simple test case below using DBMS_XMLGEN.getXML. The first row (A1) has no null values and thus tags for columns A, BEE and CEE are produced. The second row (A2) has null in column BEE and thus tags for only columns A and CEE are produced.
Is there a way to force a tag for null column BEE in the second row?
create table foo (A varchar2(10), BEE number, CEE date);
insert into foo values ('A1',1,sysdate);
insert into foo values ('A2',null,sysdate);
SELECT DBMS_XMLGEN.getXML('SELECT * FROM foo') FROM dual;
<ROWSET>
<ROW>
<A>A1</A>
<BEE>1</BEE>
<CEE>27-SEP-12</CEE>
</ROW>
<ROW>
<A>A2</A>
<CEE>27-SEP-12</CEE>
</ROW>
</ROWSET>

What's the database version? (SELECT * FROM v$version)
Could you use this instead :
SQL> select xmlserialize(document
2           xmlelement("ROWSET",
3             xmlagg(
4               xmlelement("ROW",
5                 xmlelement("A", a)
6               , xmlelement("BEE", bee)
7               , xmlelement("CEE", cee)
8               )
9             )
10           )
11         -- for display purpose only :
12         as clob indent
13         )
14 from foo
15 ;
XMLSERIALIZE(DOCUMENTXMLELEMEN
<ROWSET>
<ROW>
    <A>A1</A>
    <BEE>1</BEE>
    <CEE>2012-09-27</CEE>
</ROW>
<ROW>
    <A>A2</A>
    <BEE/>
    <CEE>2012-09-27</CEE>
</ROW>
</ROWSET>
Or,
SQL> select xmlserialize(document
2           xmlquery(
3             '(#ora:view_on_null empty #)
4             {
5               <ROWSET>{fn:collection("oradb:/DEV/FOO")}</ROWSET>
6             }'
7             returning content
8           )
9           as clob indent
10         )
11 from dual;
XMLSERIALIZE(DOCUMENTXMLQUERY(
<ROWSET>
<ROW>
    <A>A1</A>
    <BEE>1</BEE>
    <CEE>2012-09-27</CEE>
</ROW>
<ROW>
    <A>A2</A>
    <BEE/>
    <CEE>2012-09-27</CEE>
</ROW>
</ROWSET>
(where "DEV" is my test schema)
If you want to stick with DBMS_XMLGEN, you're gonna have to use PL/SQL and setNullHandling procedure.
Edited by: odie_63 on 27 sept. 2012 17:14

Replicat error: ORA-12899: value too large for column ...

Hi,
In our system Source and Target are on the same physical server and in the same Oracle instance. Just different schemes.
Tables on the target were created as 'create table ... as select * from ... source_table', so they have a similar structure. Table names are also similar.
I started replicat, it worked fine for several hours, but when I inserted Chinese symbols into the source table I got an error:
WARNING OGG-00869 Oracle GoldenGate Delivery for Oracle, OGGEX1.prm: OCI Error ORA-12899: value too large for column "MY_TARGET_SCHEMA"."TABLE1"."*FIRSTNAME*" (actual: 93, maximum: 40) (status = 12899), SQL <INSERT INTO "MY_TARGET_SCHEMA"."TABLE1" ("USERID","USERNAME","FIRSTNAME","LASTNAME",....>.
FIRSTNAME is Varchar2(40 char) field.
I suppose the problem probably is our database is running with NLS_LENGTH_SEMANTICS='CHAR'
I've double checked tables structure on the target - it's identical with the source.
I also tried to manually insert this record into the target table using 'insert into ... select * from ... ' statement - it works. The problem seems to be in the replicat.
How to fix this error?
Thanks in advance!
Oracle GoldenGate version: 11.1.1.1
Oracle Database version: 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
NLS_LANG: AMERICAN_AMERICA.AL32UTF8
NLS_LENGTH_SEMANTICS='CHAR'
Edited by: DeniK on Jun 20, 2012 11:49 PM
Edited by: DeniK on Jun 23, 2012 12:05 PM
Edited by: DeniK on Jun 25, 2012 1:55 PM

I've created the definition files and compared them. They are absolutely identical, apart from source and target schema names:
Source definition file:
Definition for table MY_SOURCE_SCHEMA.TABLE1
Record length: 1632
Syskey: 0
Columns: 30
USERID 134 11 0 0 0 1 0 8 8 8 0 0 0 0 1 0 1 3
USERNAME 64 80 12 0 0 1 0 80 80 0 0 0 0 0 1 0 0 0
FIRSTNAME 64 160 98 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
LASTNAME 64 160 264 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
PASSWORD 64 160 430 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
TITLE 64 160 596 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
Target definition file:
Definition for table MY_TAEGET_SCHEMA.TABLE1
Record length: 1632
Syskey: 0
Columns: 30
USERID 134 11 0 0 0 1 0 8 8 8 0 0 0 0 1 0 1 3
USERNAME 64 80 12 0 0 1 0 80 80 0 0 0 0 0 1 0 0 0
FIRSTNAME 64 160 98 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
LASTNAME 64 160 264 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
PASSWORD 64 160 430 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
TITLE 64 160 596 0 0 1 0 160 160 0 0 0 0 0 1 0 0 0
Edited by: DeniK on Jun 25, 2012 1:56 PM
Edited by: DeniK on Jun 25, 2012 1:57 PM

10g ORA-30625: method dispatch on NULL SELF argument is disallowed Options

I found this site for consuming a web service from oracle 9i. I'm trying
it from 10g
http://www.oracle-base.com/articles/9i/ConsumingWebServices9i.php
I have an intranet web service (.NET) I know is working as I use it from
other clients.
But am getting this error when trying the example:
Getting error: ORA-30625: method dispatch on NULL SELF argument is
disallowed.
on this line:
l_fullname := soap_api.get_return_value(p_response => l_response,
p_name => 'yyyResult',
p_namespace =>
'xmlns:ns1="urn:http://www.xxx.net/"');
which calling this in the soap_api package i downloaded and installed
from the stie:
FUNCTION get_return_value(p_response IN OUT NOCOPY t_response,
p_name IN VARCHAR2,
p_namespace IN VARCHAR2)
RETURN VARCHAR2 AS
BEGIN
RETURN p_response.doc.extract('//'||p_name||'/
child::text()',p_namespace).getstringval();
END;
My test script:
declare
l_request soap_api.t_request;
l_response soap_api.t_response;
l_fullname varchar2(50);
BEGIN
l_request := soap_api.new_request(p_method => 'ns1:yyy',
p_namespace =>
'xmlns:ns1="urn:http://www.xxx.net"');
soap_api.add_parameter(p_request => l_request,
p_name => 'userName',
p_type => 'xsd:string',
p_value => 'somestring');
l_response := soap_api.invoke(p_request => l_request,
p_url => 'http://xx.xxx.xx.xx/
securityservice/service.asmx',
p_action => 'http://www.xxx.net/
yyy');
-- failing on the below line(21)
l_fullname := soap_api.get_return_value(p_response => l_response,
p_name => 'yyyResult',
p_namespace =>
'xmlns:ns1="urn:http://www.xxx.net/"');
dbms_output.put_line(l_fullname);
end;
Thanks for any help or information.

This rather cryptic message means that the PL/SQL variable or column whose method is being invoked is null.
It's the equivilant of a null pointer exeception in java...

ORA-30625: method dispatch on NULL SELF argument is disallowed

I have streams setup with table rules.( 3 tiers). It is working fine, besides for one table when insert record, apply has error "ORA-30625: method dispatch on NULL SELF argument is disallowed". I did include all columns in supplemental log group always. Is any limit for max columns for Oracle Streams. This table has 137 columns and I created 6 supplemental log groups.
Please help

Hi,
No restrictions on number of columns in 10.2. Do you have any rule transformation or dml handler?
Check also on your bdump if the apply generated any trace file that will allow you to identify the root cause of this error.
Here is an explanation of why that error happens:
The following error occurs when passing an uninitialized object to an Object's member procedure/function:
ORA-30625: method dispatch on NULL SELF argument is disallowed
The following illustrates the error by first creating the Object specification. Notice there is one member procedure "initialize_object" that passes in SELF.
CREATE OR REPLACE TYPE role_type AS OBJECT (
role_name VARCHAR2(30),
MEMBER PROCEDURE initialize_object(SELF IN OUT role_type));
The next step is to create the body of the Object's member procedure. There is no need for any actual code as the error occurs when the NULL is attempted to be passed.
CREATE OR REPLACE TYPE BODY role_type AS
MEMBER PROCEDURE initialize_object (SELF IN OUT role_type) IS
BEGIN
NULL;
END initialize_object;
END;
The following anonymous block passes in a NULL (Uninitialized Object) to the member procedure initialize_object.
DECLARE
V_TEMP role_type;
BEGIN
ROLE_TYPE.initialize_object(V_temp);
END;
ERROR at line 1:
ORA-30625: method dispatch on NULL SELF argument is disallowed
ORA-06512: at line 4
The example has nothing to do with streams but maybe you are trying to invoke an error handler or dml handler that is causing this issue.

SOLVED !!! Import Excel Null column problem

Hi all,
Import data from Excel is fine as long as there's is no column in target table that
is Varchar2(1) and Null allowable.
If there's the above type, the DML for the insert always replace the null column with
'NULL' resulting in SQL Error:ORA-12899: value too large for column (actual: 4, maximum: 1).
Had check preferences setup and could not find option relating to this.
Btw, am using Verion 1.1.1.25, on Windows XP, 10GR2 database also on same PC.
Had use the Export Data to export above table's data and
had checked the XLS file in Excel and the null column is indeed display as blank and use the Import Data-Excel, to import same XLS file back to database.
Pls advise and thanks in advance.
Zack [ /pre]
Message was edited by:
Zack.L
SOLVED !!! Using Version 1.1.3 Build Main-27.66
A MILLION THANKS !!!
Best Regards
Zack
Message was edited by:
Zack.L

Hi Barry,
   That's GREAT NEWS !!!.
   Yes, in the Excel sheet, the column is null and when
doing import, the SQL command use 'NULL' instead of NULL
resulting in the string 'NULL' to be inserted.
   So, i just update current release 1.1.2.25 to 1.1.2.25.79 ?
Thanks for the great work.
Zack
Message was edited by:
        Zack.L

How to make sqlldr handle mutliple NULL columns

I am using TRAILING NULL COLS in the control file to tell SQL*Loader to treat any relatively positioned columns that are not present in the record as null columns.
But the problem is occuring when there are multiple NULL columns one after the other. In that case SQL*Loader is treating continuous null column as 1 NULL column and then I am getting following message:
ORA-12899: value too large for column
And reason being wrong value is being tried to inserted in wrong column.
How can I stop SQL*Loader from combining multiple and continuous NULL columns and basically to make it treat each of them separately.

I have been able to take care of the issue using the NULLIF attribute. Here is how my control file looks now:
LOAD DATA
INFILE 'tab_sal_ldr.tab'
INSERT INTO TABLE SQL_TAB_LDR
REPLACE
FIELDS TERMINATED BY "\t"
TRAILING NULLCOLS
(COL1 CHAR,
COL2 INTEGER EXTERNAL NULLIF COL2=BLANKS,
COL3 FLOAT EXTERNAL NULLIF COL3=BLANKS,
COL4 DATE EXTERNAL NULLIF COL4=BLANKS)
However it is rejecting the rows having all NULLs. Is it possible to upload rows having all column values as NULL.
Following is the snippet from the log:
2 Rows not loaded because all fields were null.
Edited by: Parag Kalra on May 4, 2010 3:31 AM

ORA-24347: Warning of a NULL column in an aggregate function

Similar Messages

Maybe you are looking for