"The man who makes no mistakes does not usually make anything. ~Edward Phelps"
if you do end up with duplicates, then here is how to solve the problem. This SQL will delete the first occurrence of the duplicate entries.
To change it so that it deletes the last entry, you would replace Min with Max.
DELETE FROM table_with_duplicates
WHERE id IN (
SELECT bad_rows.id from table_with_duplicates AS bad_rows
INNER JOIN (
SELECT column_1, column_2, Min(id) AS min_id FROM table_with_duplicates
GROUP BY column_1, column_2 HAVING COUNT (*) > 1
) AS good_rows
ON good_rows.column_1 = bad_rows.column_1
AND good_rows.column_2 = bad_rows.column_2 AND good_rows.min_id <> bad_rows.id