Data of column datatype CLOB is moved to other columns of the same table

Question

Hi all,
I have an issue with the tables having a CLOB datatype field.
When executing a simple query on a table with a column of type CLOB it returns error [POL-2403] value too large for column.
SQL> desc od_stock_nbcst_notes;
Name Null? Type
OD_STOCKID N NUMBER
NBC_SERVICETYPE N VARCHAR(40)
LANGUAGECODE N VARCHAR(8)
AU_USERIDINS Y NUMBER
INSERTDATE Y DATE
AU_USERIDUPD Y NUMBER
MODIFYDATE Y DATE
VERSION Y SMALLINT(4)
DBUSERINS Y VARCHAR(120)
DBUSERUPD Y VARCHAR(120)
TEXT Y CLOB(2000000000)
NBC_PROVIDERCODE N VARCHAR(40)
SQL> select * from od_stock_nbcst_notes;
[POL-2403] value too large for column
Checking deeply, some of the rows have got the data of the CLOB column moved in another column of the table.
When doing select length(nbc_providercode) the length is bigger than the datatype of the field (varchar(40)).
When doing substr(nbc_providercode,1,40) to see the content of the field, a portion of the Clob data is retrieved.
SQL> select max(length(nbc_providercode)) from od_stock_nbcst_notes;
MAX(LENGTH(NBC_PROVIDERCODE))
162
Choosing one random record, this is the stored information.
SQL> select length(nbc_providerCode), text from od_stock_nbcst_notes where length(nbc_providerCode)=52;
LENGTH(NBC_PROVIDERCODE) | TEXT
-------------------------+-----------
52 | poucos me
SQL> select nbc_providerCode from od_stock_nbcst_notes where length(nbc_providerCode)=52;
[POL-2403] value too large for column
SQL> select substr(nbc_providercode,1,40) from od_stock_nbcst_notes where length(nbc_providercode)=52 ;
SUBSTR(NBC_PROVIDERCODE
Aproveite e deixe o seu carro no parque
The content of the field is part of the content of the field text (datatype CLOB, containts an XML)!!!
The right content of the field must be 'MTS' (retrieved from Central DB).
The CLOB is being inserted into the Central DB, not into the Client ODB. Data is synchronized from CDB to ODB and the data is reaching the client in a wrong way.
The issue can be recreated all the time in the same DB, but between different users the "corrupted" records are different.
Any idea?

Answer 1

939569 wrote:
Hello,
I am using Oracle 11.2, I would like to use SQL to update one column based on values of other rows at the same table. Here are the details:
create table TB_test (myId number(4), crtTs date, updTs date);
insert into tb_test(1, to_date('20110101', 'yyyymmdd'), null);
insert into tb_test(1, to_date('20110201', 'yyyymmdd'), null);
insert into tb_test(1, to_date('20110301', 'yyyymmdd'), null);
insert into tb_test(2, to_date('20110901', 'yyyymmdd'), null);
insert into tb_test(2, to_date('20110902', 'yyyymmdd'), null);
After running the SQL, I would like have the following result:
1, 20110101, 20110201
1, 20110201, 20110301
1, 20110301, null
2, 20110901, 20110902
2, 20110902, null
Thanks for your suggestion.How do I ask a question on the forums?
SQL and PL/SQL FAQ