Re: Frequently updated tables

Поиск
Список
Период
Сортировка
От pgsql@mohawksoft.com
Тема Re: Frequently updated tables
Дата
Msg-id 17403.24.91.171.78.1086783688.squirrel@mail.mohawksoft.com
обсуждение исходный текст
Ответ на Re: Frequently updated tables  (Mark Kirkwood <markir@coretech.co.nz>)
Ответы Re: Frequently updated tables  (Christopher Kings-Lynne <chriskl@familyhealth.com.au>)
Список pgsql-hackers
>
> pgsql@mohawksoft.com wrote:
>
>>The best phrasing would be "the accumulating overhead of deletes and
>>updates."
>>
>>Yes.
>>
>>
>
> Are you using 7.3?
>
> I am asking because in 7.3 high update / delete tables could suffer
> (index and toast) bloat that was untamable via (lazy) VACUUM and FSM.
> I believe this is fixed in 7.4, so it should be possible to achieve on
> disk size control of tables / indexes by configuring FSM and (lazy)
> VACUUM. Did you find this not to be the case?
>
Interesting, the company is usng 7.3.4. One single row summary table got
up to 2 million dead rows. A select from that single row took a quarter of
a second. A regular vacuum did not fix it, only a vacuum full did.
However, when the test was re-run with constant vacuums, it did not get
out of hand.

My concern is performance, and yes, for inserts PostgreSQL is great. For
data that is constantly being updated, PostgreSQL is a bit weak. Think
about a table with a few million rows that needs to be updated a few
thousand times a minute.

I love PG, I've been using it since version 6x, and it has gotten
fantastic over the years, and in many cases, I would choose it over
Oracle, but for systems that need frequent updates, I have a lot of
concerns.


В списке pgsql-hackers по дате отправления:

Предыдущее
От: Andrew Dunstan
Дата:
Сообщение: Re: Question regarding dynamic_library_path
Следующее
От: "Thomas Hallgren"
Дата:
Сообщение: Re: Question regarding dynamic_library_path