Re: Re: GiST optimizing memmoves in gistplacetopage for fixed-size updates [PoC]
От | Andrew Borodin |
---|---|
Тема | Re: Re: GiST optimizing memmoves in gistplacetopage for fixed-size updates [PoC] |
Дата | |
Msg-id | CAJEAwVHhaieEK3A7PsGunDJMfH6vLPVKKwZAjSuK4zH-cFhMXg@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Re: GiST optimizing memmoves in gistplacetopage for fixed-size updates [PoC] (Alvaro Herrera <alvherre@2ndquadrant.com>) |
Список | pgsql-hackers |
>Can you please patch BRIN to use this new function too? On my machine replacement of both BRIN update cases (see https://github.com/x4m/pggistopt/commit/a6d301ff79104b977619339d53aebf748045418a ) showed no performance changes on folowing update query (6 seconds of updates avg): create table dataTable(x int, y int); insert into dataTable(x,y) select x,y from generate_series(1,1e3,1) x,generate_series(1,1e3,1) y; create index idx on dataTable using brin(x,y); update datatable set x = random()*1024, y = random()*1024; https://gist.github.com/x4m/7e69fd924b9ffd2fdc9c6100e741171d Probably I was looking in a wrong place. I do not see other cases when PageIndexTupleOverwrite can improve performance of BRIN. Though I'll make PageIndexTupleOverwrite BRIN-compatible in forthcoming patch version: BRIN tuples have no length in header and it must be taken as a parameter. Just as the PageAddItem do. Best regards, Andrey Borodin, Octonica & Ural Federal University.
В списке pgsql-hackers по дате отправления: