Re: Spilling hashed SetOps and aggregates to disk

Поиск
Список
Период
Сортировка
От Andres Freund
Тема Re: Spilling hashed SetOps and aggregates to disk
Дата
Msg-id 20180605171903.msx2swc6dsoljdtk@alap3.anarazel.de
обсуждение исходный текст
Ответ на Re: Spilling hashed SetOps and aggregates to disk  (David Fetter <david@fetter.org>)
Список pgsql-hackers
On 2018-06-05 19:04:11 +0200, David Fetter wrote:
> On Tue, Jun 05, 2018 at 02:56:23PM +1200, David Rowley wrote:
> > On 5 June 2018 at 06:52, Andres Freund <andres@anarazel.de> wrote:
> > > That part has gotten a bit easier since, because we have serialize
> > > / deserialize operations for aggregates these days.
> > 
> > True. Although not all built in aggregates have those defined.
> 
> Just out of curiosity, which ones don't? As of
> 3f85c62d9e825eedd1315d249ef1ad793ca78ed4, pg_aggregate has both of
> those as NOT NULL.

That doesn't mean much. We commonly store 0 / InvalidOid for such things
in the catalog. Allows us to still map the tuples to structs (due to
fixed width).

Greetings,

Andres Freund


В списке pgsql-hackers по дате отправления:

Предыдущее
От: "MauMau"
Дата:
Сообщение: Re: I'd like to discuss scaleout at PGCon
Следующее
От: David Fetter
Дата:
Сообщение: [PATCH] Trim trailing whitespace in vim and emacs