Does your data have duplicate key values? If so, remove the duplicates.
You could build a non-unique index on the primary key columns, and a mirror table (same cols etc), but an additional col - unique_val char(13) for bit data
insert into mirror select *, generate_unique() from original where key_col in (select key_col from original group by key_col having count(*) > 1)
delete from original where key_col in (select key_col from mirror)
delete from mirror m1 where exists (select 1 from mirror m2 where m1.unique_val > m2.unique_val) << some versions of DB2 don't support this, you may need another copy of mirror to make it work.
insert into original select col1, col2, ... <without unique_val> from mirror
It appears you are running DB2 for MVS/OS390.zOS. Alternatively, you could rerun the load using ENFORCE CONSTRAINTS.
James Campbell