Subject: Re: logical replication
To: Marc Millas <marc.millas@mokadb.com>, pgsql-performance@postgresql.org
References: 
 <CADX_1aaqJiO1XU1OkXcX2j2KRt3ei97TVuXQiCuMWNFUpdXV_A@mail.gmail.com>
From: Andrew Dunstan <andrew@dunslane.net>
Message-ID: <d7f1bac2-8549-9d8e-956d-e635960ccaa4@dunslane.net>
Date: Fri, 21 May 2021 09:21:13 -0400
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.4.0
MIME-Version: 1.0
In-Reply-To: 
 <CADX_1aaqJiO1XU1OkXcX2j2KRt3ei97TVuXQiCuMWNFUpdXV_A@mail.gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Content-Language: en-US
Archived-At: 
 <https://www.postgresql.org/message-id/d7f1bac2-8549-9d8e-956d-e635960ccaa4%40dunslane.net>
Precedence: bulk


On 5/21/21 8:41 AM, Marc Millas wrote:
> Hi,
>
> my POC in postgres 12.(important ?)
>
> if I setup 2 postgres clusters, and create a publication in one and a
> subscription in the other,
> and do on the pub an update which does not change the data (updating
> an existing record with same data) then this (useless) update go
> through replication.(ie consumes network ressource)
>
> what are ways to avoid this ?
> (I thought of a trigger to not execute the useless update, but I
> dont see how to do this)
> any ideas ?
>
> thanks
> PS: remarks about the meaning of this are off topic, thanks
>
>

Postgres provides exactly such a trigger. See
https://www.postgresql.org/docs/12/functions-trigger.html


cheers


andrew

--
Andrew Dunstan
EDB: https://www.enterprisedb.com