Thread: [Fwd: Bug#255208: postgresql - ignores SIGPIPE]

This seems a reasonable request.  Is it possible?

-----Forwarded Message-----
From: Bastian Blank <waldi@debian.org>
To: submit@bugs.debian.org
Subject: Bug#255208: postgresql - ignores SIGPIPE
Date: Sat, 19 Jun 2004 16:52:04 +0200

Package: postgresql
Version: 7.4.3-1
Severity: important

The postgres daemon ignores SIGPIPE. This makes it impossible to cancel
queries if the client is not able to issue a cancel command, like the
perl and python bindings.

The code contains a signal handler for SIGINT which cancels the actual
query, so I see no problem to have SIGPIPE behave the same way.

Bastian
--
Oliver Elphick                                          olly@lfix.co.uk
Isle of Wight                              http://www.lfix.co.uk/oliver
GPG: 1024D/A54310EA  92C8 39E7 280E 3631 3F0E  1EC0 5664 7A2F A543 10EA
                 ========================================
     "Therefore being justified by faith, we have peace with
      God through our Lord Jesus Christ."       Romans 5:1

Oliver Elphick <olly@lfix.co.uk> writes:
> The postgres daemon ignores SIGPIPE.

This is not a bug; it's intentional.

            regards, tom lane

Hi Tom, hi Bastian!

On 2004-06-20  7:49 -0400, Tom Lane wrote:
> Oliver Elphick <olly@lfix.co.uk> writes:
> > The postgres daemon ignores SIGPIPE.
>=20
> This is not a bug; it's intentional.

The original submitter (Bastian Blank) commented on this:

> The reason is not documented and the daemon is able to abort queries in
> this state.
>=20
> It makes systems unusable by stuck server processes which aren't
> abortable without admin intervention.
>=20
> Bastian

Thanks and have a nice day,

Martin
--=20
Martin Pitt                 Debian GNU/Linux Developer
martin@piware.de                      mpitt@debian.org
http://www.piware.de             http://www.debian.org