Skip site navigation (1) Skip section navigation (2)

Re: Postgres && Swap

From: "Scott Marlowe" <scott(dot)marlowe(at)gmail(dot)com>
To: steve(at)outtalimits(dot)com(dot)au
Cc: pgsql-admin(at)postgresql(dot)org
Subject: Re: Postgres && Swap
Date: 2008-09-18 00:40:19
Message-ID: dcc563d10809171740uef9013clee242c224ed4ced5@mail.gmail.com (view raw or flat)
Thread:
Lists: pgsql-admin
Well, there's plenty of negatives to not doing it.  Like postgresql's
large shared_buffers getting swapped out to make more space for disk
buffers.  So, pgsql goes to grab data from shared buffers and has to
wait for the OS to swap them back in.

On Wed, Sep 17, 2008 at 4:28 PM,  <steve(at)outtalimits(dot)com(dot)au> wrote:
> Is there any negative effects by doing this?
>
> The swappiness is sitting on the default 60 at the moment.
>
> On Wed, 17 Sep 2008 00:46:13 -0600, "Scott Marlowe"
> <scott(dot)marlowe(at)gmail(dot)com> wrote:
>> On Tue, Sep 16, 2008 at 11:30 PM,  <steve(at)outtalimits(dot)com(dot)au> wrote:
>>> Hi All
>>>
>>>
>>> I recently made a change to my Postgres Server and upped the
>> max_fsm_page
>>> size to 60000
>>> Since then, Postgres has been using about 30-80MB of swap space.
>>>
>>> This box has 4GB of RAM. All up Postgres has not been allocated no more
>>> than 3G
>>>
>>> Is this swapping something to be worried about?
>>
>> I recommend  turning swappiness down in linux that should stop pg from
>> getting prematurely swapped out like that.
>

In response to

pgsql-admin by date

Next:From: Julius TuskenisDate: 2008-09-18 09:19:13
Subject: CVS for postgresql
Previous:From: steveDate: 2008-09-17 22:28:36
Subject: Re: Postgres && Swap

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group