Date: Thu, 24 Jul 2025 15:48:37 +0900 (JST)
Message-Id: <20250724.154837.765982152313889926.ishii@postgresql.org>
To: emond.papegaaij@gmail.com
Cc: pgpool-general@lists.postgresql.org
Subject: Re: FATAL: simple query "BEGIN" arrived before ending an extended
 query message
From: Tatsuo Ishii <ishii@postgresql.org>
In-Reply-To: 
 <CAGXsc+Z6WFxbORtXjdhHXoDLZ_9e_5BSFZK1ju9As3-8dA82OA@mail.gmail.com>
References: <20250723.170320.645068218702820010.ishii@postgresql.org>
	<20250724.103607.89592995648988686.ishii@postgresql.org>
	<CAGXsc+Z6WFxbORtXjdhHXoDLZ_9e_5BSFZK1ju9As3-8dA82OA@mail.gmail.com>
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Archived-At: 
 <https://www.postgresql.org/message-id/20250724.154837.765982152313889926.ishii%40postgresql.org>
Precedence: bulk

> Hi,
> 
> There's been some recent development on the side of pgjdbc as well:
> https://github.com/pgjdbc/pgjdbc/pull/3728
> 
> We are currently testing that fix. With that patch, pgjdbc will always send
> a sync first when in extended mode before sending a simple query.

Ok. That's good.

> Do you
> think it's still a good idea to support mixing simple and extended mode in
> pgpool? It's difficult for me to assess the risk of your proposed patch.

I agree with your concern. Mixing simple and extended mode is a new
code path in pgpool. Probably mixing simple and extended mode is also
new code path in PostgreSQL backend. We can't ignore both risks.

> Maybe it's better to just not allow it if that makes it easier to implement?

I agree. I withdraw my patch.

> Ps. Sorry for the bad quoting.

No problem at all.

> I'm writing this on my phone, as I'm still
> on vacation.

Sorry for bothering you vacation. Have a nice day!

Best regards,
--
Tatsuo Ishii
SRA OSS K.K.
English: http://www.sraoss.co.jp/index_en/
Japanese:http://www.sraoss.co.jp