From: Kevin Grittner <kgrittn@ymail.com>
To: Евгений Селявка <evg.selyavka@gmail.com>
To: pgsql-admin@postgresql.org <pgsql-admin@postgresql.org>
Subject: Re: postgresql 9.1 out of memory
Date: Wed, 30 Oct 2013 08:14:09 -0700 (PDT)
Message-ID: <1383146049.700.YahooMailNeo@web162906.mail.bf1.yahoo.com> (raw)
In-Reply-To: <CAKPhvNY+LJV1MDTW=1KHX+igoUyyPVW8_V1-5+T18qBQ3CJGrw@mail.gmail.com>
References: <CAKPhvNY+LJV1MDTW=1KHX+igoUyyPVW8_V1-5+T18qBQ3CJGrw@mail.gmail.com>
List-Unsubscribe: <mailto:majordomo@postgresql.org?body=unsub%20pgsql-admin>
Евгений Селявка <evg.selyavka@gmail.com> wrote:
> Server HW:
> 32GB RAM
> Before server crash i have this parameters in config:
>
> max_connections = 350
> work_mem = 2GB
> After crash i change this parameters:
> work_mem = 1GB
A good "rule of thumb" is to allow for one work_mem allocation per
connection for a reasonable estimation of peak memory usage for
this purpose. (This may need to be adjusted based on workload, but
it's a reasonable place to start.) So you have adjusted likely
peaks down from 700GB to 350GB.
I usually start with work_mem set to RAM * 25% / max_connections.
In your case that works out to 23MB. If you are able to use a
connection pooler to reducee max_connections, that will allow you
to boost work_mem.
--
Kevin Grittner
EDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
--
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-admin@postgresql.org
Cc: kgrittn@ymail.com, evg.selyavka@gmail.com
Subject: Re: postgresql 9.1 out of memory
In-Reply-To: <1383146049.700.YahooMailNeo@web162906.mail.bf1.yahoo.com>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox