MIME-Version: 1.0
References: <LO0P265MB2700E0D30CC9A427956FC121F9F02@LO0P265MB2700.GBRP265.PROD.OUTLOOK.COM>
In-Reply-To: <LO0P265MB2700E0D30CC9A427956FC121F9F02@LO0P265MB2700.GBRP265.PROD.OUTLOOK.COM>
From: Francisco Olarte <folarte@peoplecall.com>
Date: Tue, 28 May 2024 19:14:49 +0200
Message-ID: <CA+bJJbyzFdr1i1Yarqt=JnsN0WUg+veRvYZC1+kaYnWFGjiYRw@mail.gmail.com>
Subject: Re: Memory issues with PostgreSQL 15
To: =?UTF-8?Q?Christian_Schr=C3=B6der?= <christian.schroeder@wsd.com>
Cc: "pgsql-general@lists.postgresql.org" <pgsql-general@lists.postgresql.org>, Eric Wong <eric.wong@wsd.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://www.postgresql.org/message-id/CA%2BbJJbyzFdr1i1Yarqt%3DJnsN0WUg%2BveRvYZC1%2BkaYnWFGjiYRw%40mail.gmail.com>
Precedence: bulk

Hi Christian:

On Tue, 28 May 2024 at 18:40, Christian Schr=C3=B6der
<christian.schroeder@wsd.com> wrote:

> <2024-05-21 11:34:46 CEST - mailprocessor> ERROR:  could not resize share=
d memory segment "/PostgreSQL.2448337832" to 182656 bytes: No space left on=
 device

This hints at some shm function getting an ENOSPC: Coupled with...

> I thought this could all be related to our "shared_buffers" setting, so I=
 increased it to 8 GB. This almost immediately (after a few minutes) gave m=
e these errors:

A faster fail when increasing it I would start by checking your IPC
shared memory limits are ok, especially if you upgraded something in
the OS when going from 9 to 15, which seems likely.

IIRC in linux you can read them in /proc/sys/kernel/shm*, and they
were configured via sysctl.

Francisco Olarte.