MIME-Version: 1.0
References: 
 <CAA5-nLCdhiJW=zEBexw=NeDGirNGdKVWb-TEhRALM6+DCDRs6w@mail.gmail.com>
 <44166282b3516f33696f025ad9f5271244f78107.camel@cybertec.at>
 <CAA5-nLB9x9SW4KoPcukxAMj9N=wARJLHEBfkkGW+0Hh2FarnPA@mail.gmail.com>
In-Reply-To: 
 <CAA5-nLB9x9SW4KoPcukxAMj9N=wARJLHEBfkkGW+0Hh2FarnPA@mail.gmail.com>
From: Fabrice Chapuis <fabrice636861@gmail.com>
Date: Mon, 2 Mar 2026 17:13:22 +0100
Message-ID: 
 <CAA5-nLBjwDeXTbsMxL-fwbfQ8asoCG0uG6hCMzPt2MLysdwFwQ@mail.gmail.com>
Subject: Re: problem during postgres restore: max_connections = 100 is a lower
 setting than on the primary server, where its value was 350
To: Pgsql-admin <pgsql-admin@lists.postgresql.org>,
	Laurenz Albe <laurenz.albe@cybertec.at>
Content-Type: multipart/alternative; boundary="00000000000047aec5064c0cde0e"
Archived-At: 
 <https://www.postgresql.org/message-id/CAA5-nLBjwDeXTbsMxL-fwbfQ8asoCG0uG6hCMzPt2MLysdwFwQ%40mail.gmail.com>
Precedence: bulk

--00000000000047aec5064c0cde0e
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Thanks for answering Laurenz,

I think the value of max_connections must be aligned with primary because
the goal is to maintain this sizing in case of a failover and promotion of
the standby.
Why to be conservative with `max_connection` and not with shared buffer?
If we're performing recovery on a machine with significantly fewer CPU and
RAM resources than the original server, lowering these parameters could be
an option because they reserve memory at starting.

These are the questions I've been asking to myself.

On Mon, Mar 2, 2026 at 4:20=E2=80=AFPM Fabrice Chapuis <fabrice636861@gmail=
.com>
wrote:

> Thanks for answering Laurenz,
>
> I think the value of max_connections must be aligned with primary because
> the goal is to maintain this sizing in case of a failover and promotion o=
f
> the standby.
> Why to be conservative with `max_connection` and not with shared buffer?
> If we're performing recovery on a machine with significantly fewer CPU an=
d
> RAM resources than the original server, lowering these parameters could b=
e
> an option because they reserve memory at starting.
>
> These are the questions I've been asking to myself.
>
> On Mon, Mar 2, 2026 at 3:36=E2=80=AFPM Laurenz Albe <laurenz.albe@cyberte=
c.at>
> wrote:
>
>> On Mon, 2026-03-02 at 11:27 +0100, Fabrice Chapuis wrote:
>> > I got this error from restoring a backup on a recovery machine.
>> > I got error about the max_connections that I changed from 350 to 100.
>> > in the control file the value is 350
>> > /usr/pgsql-17/bin/pg_controldata -D /pgdata/pgsql/17/main/ | grep
>> max_conn
>> > max_connections setting:              350
>> > That is the value registered from the primary. But the primary is not
>> intended to be connected during this restore.
>> > primary_conninfo is desactivated.
>> >
>> > Is there a trick to change decrease this value and start the instance.
>>
>> Why bother?
>>
>> Leave max_connections at 350.
>>
>> Yours,
>> Laurenz Albe
>>
>

--00000000000047aec5064c0cde0e
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Thanks for answering Laurenz,<div><br>I think the value of=
 max_connections must be aligned with primary because the goal is to mainta=
in this sizing in case of a failover and promotion of the standby.<br>Why t=
o be conservative with `max_connection` and not with shared buffer?<br>If w=
e&#39;re performing recovery on a machine with significantly fewer CPU and =
RAM resources than the original server, lowering these parameters could be =
an option because they reserve memory at starting.<br><br>These are the que=
stions I&#39;ve been asking to myself.</div></div><br><div class=3D"gmail_q=
uote gmail_quote_container"><div dir=3D"ltr" class=3D"gmail_attr">On Mon, M=
ar 2, 2026 at 4:20=E2=80=AFPM Fabrice Chapuis &lt;<a href=3D"mailto:fabrice=
636861@gmail.com">fabrice636861@gmail.com</a>&gt; wrote:<br></div><blockquo=
te class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px =
solid rgb(204,204,204);padding-left:1ex"><div dir=3D"ltr">Thanks for answer=
ing Laurenz,<div><br>I think the value of max_connections must be aligned w=
ith primary because the goal is to maintain this sizing in case of a failov=
er and promotion of the standby.<br>Why to be conservative with `max_connec=
tion` and not with shared buffer? <br>If we&#39;re performing recovery on a=
 machine with significantly fewer CPU and RAM resources than the original s=
erver, lowering these parameters could be an option because they reserve me=
mory at starting.<br><br>These are the questions I&#39;ve been asking to my=
self.</div></div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"g=
mail_attr">On Mon, Mar 2, 2026 at 3:36=E2=80=AFPM Laurenz Albe &lt;<a href=
=3D"mailto:laurenz.albe@cybertec.at" target=3D"_blank">laurenz.albe@cyberte=
c.at</a>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"mar=
gin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1=
ex">On Mon, 2026-03-02 at 11:27 +0100, Fabrice Chapuis wrote:<br>
&gt; I got this error from restoring a backup on a recovery machine.<br>
&gt; I got error about the max_connections that I changed from 350 to 100.<=
br>
&gt; in the control file the value is 350<br>
&gt; /usr/pgsql-17/bin/pg_controldata -D /pgdata/pgsql/17/main/ | grep max_=
conn<br>
&gt; max_connections setting: =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0350<br>
&gt; That is the value registered from the primary. But the primary is not =
intended to be connected during this restore.<br>
&gt; primary_conninfo is desactivated.<br>
&gt; <br>
&gt; Is there a trick to change decrease this value and start the instance.=
<br>
<br>
Why bother?<br>
<br>
Leave max_connections at 350.<br>
<br>
Yours,<br>
Laurenz Albe<br>
</blockquote></div>
</blockquote></div>

--00000000000047aec5064c0cde0e--