Re: crash-safe visibility map, take three

Поиск
Список
Период
Сортировка
От Robert Haas
Тема Re: crash-safe visibility map, take three
Дата
Msg-id AANLkTi=7St=B=MONtpwD7eTQe9o2Vzd2j5Vfs0wnmj0b@mail.gmail.com
обсуждение исходный текст
Ответ на Re: crash-safe visibility map, take three  (Tom Lane <tgl@sss.pgh.pa.us>)
Ответы Re: crash-safe visibility map, take three  (Robert Haas <robertmhaas@gmail.com>)
Re: crash-safe visibility map, take three  (Heikki Linnakangas <heikki.linnakangas@enterprisedb.com>)
Список pgsql-hackers
On Tue, Nov 30, 2010 at 11:16 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> How much is "quite a lot"?  Do we have any real reason to think that
> this solution is unacceptable performance-wise?

Well, let's imagine a 1GB insert-only table.  It has 128K pages.  If
you XLOG setting the bit on each page, you'll need to write 128K WAL
records, each containing a 12-byte relfilenode and a 4-byte block
offset, for a total of 16 bytes of WAL per page, thus 2MB of WAL.

But you did just dirty a gigabyte of data.

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


В списке pgsql-hackers по дате отправления:

Предыдущее
От: Tom Lane
Дата:
Сообщение: Re: crash-safe visibility map, take three
Следующее
От: Robert Haas
Дата:
Сообщение: Re: [GENERAL] column-level update privs + lock table