MIME-Version: 1.0
References: 
 <CAO6_Xqo3co3BuUVEVzkaBVw9LidBgeeQ_2hfxeLMQcXwovB3GQ@mail.gmail.com>
 <CAO6_XqrZEREa5d+dyjahX6bteBhoN=8Jid-3a4f6Q35sWrv9eg@mail.gmail.com>
 <CAHGQGwHFKF+x4E+SqedMCnmLCitxjTUUtSyL_+mMeuq-GbEt6w@mail.gmail.com>
 <CAO6_Xqp+ADb6KZVWLMALu3xmwVUEO8S1EiCnp38mG6BrHrEnuA@mail.gmail.com>
 <CAO6_XqqKDV+AuP=Gf4kRKPqzyYTsOyGd3LE8Jqkwi7EMPJpbhA@mail.gmail.com>
 <CAHGQGwHc5yH4Nxp59KXJP0kAr61j3W7QeSKT2HxVjZa3OrLzmg@mail.gmail.com>
 <CAO6_Xqq1h6kggb1o206rgouPS0H5jnjahzZ0We-9ggnBjB2JsA@mail.gmail.com>
 <CAHGQGwFJnNUOMiW9wR-2WjSKzzj0wV8p55J8bnJ6mik=z0oFPQ@mail.gmail.com>
 <2631a3c3-5e60-4a1a-9e20-377024322602@gmail.com>
 <CAO6_Xqq73TPa3M6nQ7RqRhKkcphy1JX7aNGTYy-x_Sn+6a8Z_Q@mail.gmail.com>
In-Reply-To: 
 <CAO6_Xqq73TPa3M6nQ7RqRhKkcphy1JX7aNGTYy-x_Sn+6a8Z_Q@mail.gmail.com>
From: Fujii Masao <masao.fujii@gmail.com>
Date: Fri, 6 Mar 2026 08:46:15 +0900
Message-ID: 
 <CAHGQGwGvnpN=2bo+F7H90YLFcx9=SazwLkcx+0gEcrbQy5NVZg@mail.gmail.com>
Subject: Re: Shutdown indefinitely stuck due to unflushed FPI_FOR_HINT record
To: Anthonin Bonnefoy <anthonin.bonnefoy@datadoghq.com>
Cc: Alexander Lakhin <exclusion@gmail.com>,
 PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Content-Type: multipart/mixed; boundary="000000000000998d22064c4f8b11"
Archived-At: 
 <https://www.postgresql.org/message-id/CAHGQGwGvnpN%3D2bo%2BF7H90YLFcx9%3DSazwLkcx%2B0gEcrbQy5NVZg%40mail.gmail.com>
Precedence: bulk

--000000000000998d22064c4f8b11
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Mar 5, 2026 at 5:40=E2=80=AFPM Anthonin Bonnefoy
<anthonin.bonnefoy@datadoghq.com> wrote:
> So it was relying on GetInsertRecPtr() instead of
> GetXLogInsertRecPtr(). As mentioned in the thread, GetInsertRecPtr()
> only returns the position of the last full xlog page, meaning it
> doesn't fix the issue we have where the last partial page contains a
> continuation record.
>
> Testing the XLogFlush(GetInsertRecPtr()) patch with my script, I still
> get the shutdown stuck issue.
>
> Using GetXLogInsertRecPtr() is required to make sure the last partial
> page is correctly flushed.

Since GetXLogInsertRecPtr() returns a bogus LSN and XLogFlush() does
almost nothing during recovery, I added a !RecoveryInProgress() check
as follows. I've attached the latest version of the patch and updated
the commit message.

- if (got_STOPPING)
- XLogBackgroundFlush();
+ if (got_STOPPING && !RecoveryInProgress())
+ XLogFlush(GetXLogInsertRecPtr());

Regards,

--=20
Fujii Masao

--000000000000998d22064c4f8b11
Content-Type: application/octet-stream;
	name="v6-0001-Fix-publisher-shutdown-hang-caused-by-logical-wal.patch"
Content-Disposition: attachment;
	filename="v6-0001-Fix-publisher-shutdown-hang-caused-by-logical-wal.patch"
Content-Transfer-Encoding: base64
Content-ID: <f_mme3usjo0>
X-Attachment-Id: f_mme3usjo0
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--000000000000998d22064c4f8b11--