Re: some queries on standby preventing replication updates

Поиск
Список
Период
Сортировка
От Joe Van Dyk
Тема Re: some queries on standby preventing replication updates
Дата
Msg-id CACfv+pJ9G4cUvwre5MRqGj+romQZmQ29GK6Xf-AenOgaXcH=kQ@mail.gmail.com
обсуждение исходный текст
Ответ на Re: some queries on standby preventing replication updates  (Emanuel Calvo <emanuel.calvo@2ndquadrant.com>)
Ответы Re: some queries on standby preventing replication updates  (Emanuel Calvo <emanuel.calvo@2ndquadrant.com>)
Re: some queries on standby preventing replication updates  (David G Johnston <david.g.johnston@gmail.com>)
Re: some queries on standby preventing replication updates  (Torsten Förtsch <torsten.foertsch@gmx.net>)
Список pgsql-general
On Mon, Oct 27, 2014 at 6:22 AM, Emanuel Calvo <emanuel.calvo@2ndquadrant.com> wrote:

El 23/10/14 a las 17:40, Joe Van Dyk escibió:
> Hi,
>
> I have a master and a slave database.
>
> I've got hot_standby_feedback turned on,
> max_standby_streaming_delay=-1. I've configured the master and slave
> to keep a few days of WALs around.
>
> I've noticed that when some large queries are run on the standby
> machine (ones that take more than a minute or so), replication updates
> are paused. Is there a way to fix this?
>
You may need to set a value on max_standby_streaming_delay, which
controls the time
before cancelling the standby queries when a conflict occurs on a
wal-records-about-to-be-applied.

Source:
http://www.postgresql.org/docs/9.3/static/runtime-config-replication.html

I'm using -1 for that option, would using something different be better? 

В списке pgsql-general по дате отправления:

Предыдущее
От: Joe Van Dyk
Дата:
Сообщение: Re: pg killed by oom-killer, "invalid contrecord length 2190 at A6C/331AAA90" on slaves
Следующее
От: Emanuel Calvo
Дата:
Сообщение: Re: some queries on standby preventing replication updates