Feedback-ID: i0c21471d:Fastmail
User-Agent: Cyrus-JMAP/3.11.0-alpha0-480-g515a2f54a-fm-20240515.001-g515a2f54
MIME-Version: 1.0
Message-Id: <009dbc5c-d229-4180-b425-632b681cba1d@app.fastmail.com>
In-Reply-To: <171606508791.686.3944833439498473045@wrigleys.postgresql.org>
References: <171606508791.686.3944833439498473045@wrigleys.postgresql.org>
Date: Mon, 20 May 2024 23:16:17 -0300
From: "Euler Taveira" <euler@eulerto.com>
To: kdg.dev@gmail.com, pgsql-docs@lists.postgresql.org
Subject: Re: Unclear on Publication documentation page
Content-Type: multipart/alternative;
 boundary=0a5dfa81189c4ece8be045ce6138ff44
Archived-At: 
 <https://www.postgresql.org/message-id/009dbc5c-d229-4180-b425-632b681cba1d%40app.fastmail.com>
Precedence: bulk

--0a5dfa81189c4ece8be045ce6138ff44
Content-Type: text/plain

On Sat, May 18, 2024, at 5:44 PM, PG Doc comments form wrote:
> The page in question:
> https://www.postgresql.org/docs/current/logical-replication-publication.html
> 
> It says:
> 
> "If a table without a replica identity is added to a publication that
> replicates UPDATE or DELETE operations then subsequent UPDATE or DELETE
> operations will cause an error on the publisher. INSERT operations can
> proceed regardless of any replica identity."
> 
> The text is ambiguous as to what is meant by "the publisher" in "will cause
> an error on the publisher". I've personally never set up logical
> replication, which is why I'm reading the documentation now.
> 
> The question for me is: Does "the publisher" refer to a logical entity,
> created by the "CREATE PUBLICATION " command? And further publication in the
> publication<->subscriber model will not work anymore? Or does it refer to
> the actual database itself, and will the functionality of the database be
> impacted? As in: your live production database can now no longer process
> UPDATE and DELETE statements so long as that publication exists/so long as
> that publication is configured to publish that particular table?

The first paragraph in the referred URL says

The node where a publication is defined is referred to as publisher.

And [2] defines the logical replication model:

Logical replication uses a publish and subscribe model with one or more
subscribers subscribing to one or more publications on a publisher node.

Unless there is a documentation bug or you want to suggest changes to any part of
the documentation, use pgsql-general for questions instead.


[2] https://www.postgresql.org/docs/current/logical-replication.html


--
Euler Taveira
EDB   https://www.enterprisedb.com/

--0a5dfa81189c4ece8be045ce6138ff44
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title><style type=3D"text/css">p.Mso=
Normal,p.MsoNoSpacing{margin:0}</style></head><body><div>On Sat, May 18,=
 2024, at 5:44 PM, PG Doc comments form wrote:<br></div><blockquote type=
=3D"cite" id=3D"qt" style=3D""><div>The page in question:<br></div><div>=
<a href=3D"https://www.postgresql.org/docs/current/logical-replication-p=
ublication.html">https://www.postgresql.org/docs/current/logical-replica=
tion-publication.html</a><br></div><div><br></div><div>It says:<br></div=
><div><br></div><div>"If a table without a replica identity is added to =
a publication that<br></div><div>replicates UPDATE or DELETE operations =
then subsequent UPDATE or DELETE<br></div><div>operations will cause an =
error on the publisher. INSERT operations can<br></div><div>proceed rega=
rdless of any replica identity."<br></div><div><br></div><div>The text i=
s ambiguous as to what is meant by "the publisher" in "will cause<br></d=
iv><div>an error on the publisher". I've personally never set up logical=
<br></div><div>replication, which is why I'm reading the documentation n=
ow.<br></div><div><br></div><div>The question for me is: Does "the publi=
sher" refer to a logical entity,<br></div><div>created by the "CREATE PU=
BLICATION " command? And further publication in the<br></div><div>public=
ation&lt;-&gt;subscriber model will not work anymore? Or does it refer t=
o<br></div><div>the actual database itself, and will the functionality o=
f the database be<br></div><div>impacted? As in: your live production da=
tabase can now no longer process<br></div><div>UPDATE and DELETE stateme=
nts so long as that publication exists/so long as<br></div><div>that pub=
lication is configured to publish that particular table?<br></div></bloc=
kquote><div><br></div><div>The first paragraph in the referred URL says<=
br></div><div><br></div><div>The node where a publication is defined is =
referred to as publisher.<br></div><div><br></div><div>And [2] defines t=
he logical replication model:<br></div><div><br></div><div>Logical repli=
cation uses a publish and subscribe model with one or more<br></div><div=
>subscribers subscribing to one or more publications on a publisher node=
.<br></div><div><br></div><div>Unless there is a documentation bug or yo=
u want to suggest changes to any part of<br></div><div>the documentation=
, use pgsql-general for questions instead.<br></div><div><br></div><div>=
<br></div><div>[2]&nbsp;<a href=3D"https://www.postgresql.org/docs/curre=
nt/logical-replication.html">https://www.postgresql.org/docs/current/log=
ical-replication.html</a><br></div><div><br></div><div><br></div><div id=
=3D"sig112409080"><div class=3D"signature">--<br></div><div class=3D"sig=
nature">Euler Taveira<br></div><div class=3D"signature">EDB&nbsp;&nbsp; =
<a href=3D"https://www.enterprisedb.com/">https://www.enterprisedb.com/<=
/a><br></div></div><div><br></div></body></html>
--0a5dfa81189c4ece8be045ce6138ff44--