Message-ID: <e73f7d44c3f76455519b2286eef19b1d42d4884d.camel@cybertec.at>
Subject: Re: statement_timeout has no effect if sync standby is unavailable
From: Laurenz Albe <laurenz.albe@cybertec.at>
To: Nikhil Shetty <nikhil.dba04@gmail.com>, Pgsql-admin
	 <pgsql-admin@lists.postgresql.org>
Date: Wed, 01 Nov 2023 17:39:31 +0100
In-Reply-To: 
 <CAFpL5VzcQXxzbzkVAnRM87qVZnZX4VYh1YcWZVMqSz+jH2AErw@mail.gmail.com>
References: 
	<CAFpL5VwpZ3Ey3SL=rtXDSVo-PfYT1x6z+9z69Ssu72mCmW9_yA@mail.gmail.com>
	 <CAFpL5VzcQXxzbzkVAnRM87qVZnZX4VYh1YcWZVMqSz+jH2AErw@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.48.4 (3.48.4-1.fc38) 
MIME-Version: 1.0
Archived-At: 
 <https://www.postgresql.org/message-id/e73f7d44c3f76455519b2286eef19b1d42d4884d.camel%40cybertec.at>
Precedence: bulk

On Wed, 2023-11-01 at 15:44 +0530, Nikhil Shetty wrote:
> We have also seen this scenario when we froze the data mount point=C2=A0a=
nd run a DML
>=20
> Freeze mountpoint=C2=A0
> date;fsfreeze --freeze /postgres
> The statement is just stuck even thought statement_timeout is set
>=20
> datid        | 14175
> datname      | postgres
> pid          | 5789
> leader_pid   |=20
> usesysid     | 10
> usename      | postgres
> application_name | hang_monitor
> client_addr  | 127.0.0.1
> client_hostname |=20
> client_port  | 37440
> backend_start| 2023-11-01 10:05:51.090023+00
> xact_start   | 2023-11-01 10:05:51.167816+00
> query_start  | 2023-11-01 10:05:51.167816+00
> state_change | 2023-11-01 10:05:51.167817+00
> wait_event_type | IO
> wait_event   | WALWrite
> state        | active
> backend_xid  | 43242183
> backend_xmin |=20
> query        | update HEARTBEAT set last_updated_time =3D timezone('UTC',=
 now()) where ID =3D 1;
> backend_type | client backend

Check with "ps": perhaps the backend process is in uninterruptible sleep, w=
aiting
for I/O.

Yours,
Laurenz Albe