Re: suggestion: log_statement = sample

Поиск
Список
Период
Сортировка
От Merlin Moncure
Тема Re: suggestion: log_statement = sample
Дата
Msg-id b42b73150907211228l77c613aepeb2c8757e2826636@mail.gmail.com
обсуждение исходный текст
Ответ на Re: suggestion: log_statement = sample  ("Joshua D. Drake" <jd@commandprompt.com>)
Список pgsql-general
On Mon, Jul 20, 2009 at 1:28 PM, Joshua D. Drake<jd@commandprompt.com> wrote:
> On Mon, 2009-07-20 at 13:24 -0400, Bill Moran wrote:
>>   In actual practice, full query logging
>> is 1/50 the amount of disk I/O as the actual database activity.  If your
>> systems are so stressed that they can't handle another 2% increase, then
>> you've got bigger problems lurking.
>
> It depends on the system. I have seen even big systems take a huge hit
> by full logging due to transactional velocity.
>
> Joshua D. Drake

I've seen this too.  On high transaction load systems, log_statement
has big overhead.  I haven't tested this lately though, and I suspect
this information is version dependent (in the 'old' days, it used to
be a really big deal).

merlin

В списке pgsql-general по дате отправления:

Предыдущее
От: Merlin Moncure
Дата:
Сообщение: Re: element from an array by its index
Следующее
От: Merlin Moncure
Дата:
Сообщение: Re: ERROR: could not access status of transaction 2495690984