I'm not very familiar with those specific features in other dbms, but in mysql's case "on duplicate" and "replace" are two very different thing. The first is an

I'm not very familiar with those specific features in other dbms, but in mysql's case "on duplicate" and "replace" are two very different thing. The first is an upsert, while the other drops the old row if it exists and always make a new one. Both have different use cases.