SQLLDR direct path allows duplicates in primary key

Question

I would like to use sqlldr direct path to load millions of records in the table but direct path allows duplicates on the primary key constraints.
inserts duplicates
sqlldr deploy_ctl/deploy_ctl@dba01mdm control=ctl_test.ctl direct=true
primary key is enabled
I am not understanding the behavior that why primary key is still enabled -- (logically it should have disabled than inserted duplicates)
does not insert duplicates
sqlldr deploy_ctl/deploy_ctl@dba01mdm control=ctl_test.ctl
primary key is enabled
Please can I know if there is any work around to use direct path with primary constraints in place.

Answer 1

Dear,
Look at the following example
mhouri.world> drop table t1;
Table dropped.
mhouri.world> create table t1 (id number);
Table created.
mhouri.world> alter table t1 add constraint t1_pk primary key (id);
Table altered.
mhouri.world> select table_name, index_name, status
  2  from user_indexes
  3  where table_name = 'T1';
TABLE_NAME                     INDEX_NAME                     STATUS           
T1                             T1_PK                          VALID            
mhouri.world> select count(1) from t1;
  COUNT(1)                                                                     
         0                                                                      and here call sqlldr
C:\>sqlldr xxx/xxx@xxxx control=c.ctl bad=c.bad direct=trueAnd see what happen to the primary key index
mhouri.world> select count(1) from t1;
  COUNT(1)                                                                     
         4                                                                     
mhouri.world> select * from t1;
        ID                                                                     
         1                                                                     
         2                                                                     
         2                                                                     
         3                                                                     
mhouri.world> select table_name, index_name, status
  2  from user_indexes
  3  where table_name = 'T1';
TABLE_NAME                     INDEX_NAME                     STATUS           
T1                             T1_PK                          UNUSABLE         
mhouri.world> alter index t1_pk rebuild;
alter index t1_pk rebuild
ERROR at line 1:
ORA-01452: cannot CREATE UNIQUE INDEX; duplicate keys found The primary key index is made unusable and this is why you have duplicate records
best regards
Mohamed Houri