Remove duplicate rows in a table

Question

I have a table contains order information like below:

Order table:

As we can see from that table, each order_no has several duplicates. So

Answer 1

This should work, even in your ancient and outdated Oracle version:

delete from order_table
where rowid not in  (select min(rowid)
                     from order_table
                     group by order_no);

Answer 2

If you don't care which row you get for each order_no, perhaps the simplest solution (before Oracle 12) is:

select [whatever columns you want, probably not rn - see below]
from ( select order_table.*,
              row_number() over (partition by order_no order by null) as rn
     )
where rn = 1
;

Answer 3

Here is a query to remove semi-duplicates with precise order by differing field:

delete from message_part
where ts >= since and ts < since + 1
    and rowid not in (
        select distinct first_value(rowid) over (partition by id order by differing_field)
        from message_part where ts >= since and ts < since + 1
    );

Credits go to @a_horse_with_no_name and Oracle equivalent of Postgres' DISTINCT ON?