Re: How to speed up product code and subcode match
От | Andrus |
---|---|
Тема | Re: How to speed up product code and subcode match |
Дата | |
Msg-id | ca7c9dde-be23-46c2-227f-50a5f35e9b21@hot.ee обсуждение исходный текст |
Ответ на | Re: How to speed up product code and subcode match (Andrus <kobruleht2@hot.ee>) |
Список | pgsql-general |
Hi!
Using index
create index on toode ( split_part( toode, '/',1) )
and query
select toode.toode , n2, n3, n4
from toode, vordlusajuhinnak
where split_part( toode.toode, '/',1) = vordlusajuhinnak.toode;
reduces run time to 5 minutes.
Andrus.
23.05.2023 17:26 Andrus kirjutas:
Hi!
Separate the product code and size into two columns --- if there's somebody who really wants to see them in the above format, give them a view or generated column. Then instead of the impossible-to-optimize queries you showed, you could do something like select toode.toode , n2, n3, n4 from toode, vordlusajuhinnak where toode.toode = vordlusajuhinnak.toode;Can function index
create index on toode ( split_part( toode, '/',1) )
and query
select toode.toode , n2, n3, n4
from toode, vordlusajuhinnak
where split_part( toode.toode, '/',1) = vordlusajuhinnak.toode;
used and keeping existing table structure? Functional index should produce same speed improvement as using separate column?Andrus.
В списке pgsql-general по дате отправления: