Re: Why does WAL_DEBUG macro need to be defined by default?

Поиск
Список
Период
Сортировка
От Robert Haas
Тема Re: Why does WAL_DEBUG macro need to be defined by default?
Дата
Msg-id CA+TgmobcLLvu47HmaVY71HPcN-PkjVJvbRSg-XB2icmkJeZC3Q@mail.gmail.com
обсуждение исходный текст
Ответ на Why does WAL_DEBUG macro need to be defined by default?  (Fujii Masao <masao.fujii@gmail.com>)
Ответы Re: Why does WAL_DEBUG macro need to be defined by default?
Список pgsql-hackers
On Fri, Oct 7, 2011 at 5:19 AM, Fujii Masao <masao.fujii@gmail.com> wrote:
> I found that by default WAL_DEBUG macro has been defined in
> 9.2dev and 9.1. I'm very surprised at this. Why does WAL_DEBUG
> need to be defined by default? The performance overhead
> introduced by WAL_DEBUG is really vanishingly low?
>
> WAL_DEBUG was defined in the following commit:
> 53dbc27c62d8e1b6c5253feba04a5094cb8fe046
>
> ----------------------
>    Support unlogged tables.
>
>    The contents of an unlogged table are WAL-logged; thus, they are not
>    available on standby servers and are truncated whenever the database
>    system enters recovery.  Indexes on unlogged tables are also unlogged.
>    Unlogged GiST indexes are not currently supported.
> ----------------------

Oh, dear.  That was a mistake on my part.  :-(

The funny thing is that I've been thinking all of these months about
how convenient it is that we defined WAL_DEBUG in debug builds, not
realizing that (1) we were defining it all the time, not just in debug
builds and (2) I was the one who accidentally did that.

Sorry, all.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company


В списке pgsql-hackers по дате отправления:

Предыдущее
От: Robert Haas
Дата:
Сообщение: Re: alter table only ... drop constraint broken in HEAD
Следующее
От: Yeb Havinga
Дата:
Сообщение: Re: [REVIEW] Patch for cursor calling with named parameters