MIME-Version: 1.0
References: <CAAKRu_ZngzqnEODc7LmS1NH04Kt6Y9huSjz5pp7+DXhrjDA0gw@mail.gmail.com>
 <CAD21AoC=b09pzWNyPYuti0M2DQwy57CGF9UWWb_+v=gdWeNG=A@mail.gmail.com>
 <CAAKRu_bR2VrTUskR24GkXV15SxPaCiHYD=soOuWZcT-rGDcA8Q@mail.gmail.com>
 <CAD21AoB=EC6wzjjqDXKmNEhu_BhGg0g_Gc50yh_zvXqiDC_OSg@mail.gmail.com> <7121ac43-d6ae-33ca-8ddc-538b0c6ab492@amazon.com>
In-Reply-To: <7121ac43-d6ae-33ca-8ddc-538b0c6ab492@amazon.com>
From: John Naylor <john.naylor@enterprisedb.com>
Date: Thu, 9 Mar 2023 14:47:19 +0700
Message-ID: <CAFBsxsGLDxNus6=8LiQihrjvOrZuu7JdpC3ykDk2vzciTU6m+A@mail.gmail.com>
Subject: Re: Should vacuum process config file reload more often
To: Jim Nasby <nasbyj@amazon.com>
Cc: Masahiko Sawada <sawada.mshk@gmail.com>, Melanie Plageman <melanieplageman@gmail.com>, 
	Pg Hackers <pgsql-hackers@postgresql.org>, Andres Freund <andres@anarazel.de>
Content-Type: multipart/alternative; boundary="0000000000005ccafc05f672db06"
Archived-At: <https://www.postgresql.org/message-id/CAFBsxsGLDxNus6%3D8LiQihrjvOrZuu7JdpC3ykDk2vzciTU6m%2BA%40mail.gmail.com>
Precedence: bulk

--0000000000005ccafc05f672db06
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Mar 9, 2023 at 12:42=E2=80=AFAM Jim Nasby <nasbyj@amazon.com> wrote=
:
>
> Doesn't the dead tuple space grow as needed? Last I looked we don't
allocate up to 1GB right off the bat.

Incorrect.

> Of course, if the patch that eliminates the 1GB vacuum limit gets
committed the situation will be even worse.

If you're referring to the proposed tid store, I'd be interested in seeing
a reproducible test case with a m_w_m over 1GB where it makes things worse
than the current state of affairs.

--
John Naylor
EDB: http://www.enterprisedb.com

--0000000000005ccafc05f672db06
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><br>On Thu, Mar 9, 2023 at 12:42=E2=80=AFAM Jim Nasby &lt;=
<a href=3D"mailto:nasbyj@amazon.com" target=3D"_blank">nasbyj@amazon.com</a=
>&gt; wrote:<br>&gt;<br>&gt; Doesn&#39;t the dead tuple space grow as neede=
d? Last I looked we don&#39;t allocate up to 1GB right off the bat.<br><br>=
Incorrect.<br><br>&gt; Of course, if the patch that eliminates the 1GB vacu=
um limit gets committed the situation will be even worse.<br><br>If you&#39=
;re referring to the proposed tid store, I&#39;d be interested in seeing a =
reproducible test case with a m_w_m over 1GB where=C2=A0it makes things wor=
se than the current state of affairs.<div><br></div><div>--<br>John Naylor<=
br>EDB: <a href=3D"http://www.enterprisedb.com" target=3D"_blank">http://ww=
w.enterprisedb.com</a></div></div>

--0000000000005ccafc05f672db06--