In SQL, UNION is used to combine the results of two queries. But in many cases, UNION ALL should be used instead.

Did you know that UNION, after combining the two results, performs a deduplication operation to ensure that the result does not contain duplicate records? And that takes quite a lot of work.

In most cases, we already know from the outset that the results are mutually exclusive. For example, when selecting distinct data:

SELECT id FROM ZDOC_HEADER
UNION
SELECT id FROM ZDOC_ITEM

In this case, it makes sense to prevent the system from wasting time (and it’s a lot of time!) deduplicating something we already know has no duplicates. And in such cases, you should UNION ALL:

SELECT id FROM ZDOC_HEADER
UNION ALL
SELECT id FROM ZDOC_ITEM

Depending on the number of records, the query can easily become 20 per cent faster.

Greetings from Abapinho