From: Tom Lane <tgl@sss.pgh.pa.us>
To: Alvaro Herrera <alvherre@2ndquadrant.com>
cc: Thomas Munro <thomas.munro@gmail.com>,
        =?iso-8859-1?Q?R=E9mi?= Zara <remi_zara@mac.com>,
        Michael Paquier <michael@paquier.xyz>,
        PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Subject: Re: pgsql: Add kqueue(2) support to the WaitEventSet API.
In-reply-to: <20200224211709.GA6996@alvherre.pgsql>
References: <20200224211709.GA6996@alvherre.pgsql>
Comments: In-reply-to Alvaro Herrera <alvherre@2ndquadrant.com>
	message dated "Mon, 24 Feb 2020 18:17:09 -0300"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <5194.1582580667.1@sss.pgh.pa.us>
Date: Mon, 24 Feb 2020 16:44:27 -0500
Message-ID: <5195.1582580667@sss.pgh.pa.us>
Precedence: bulk

Alvaro Herrera <alvherre@2ndquadrant.com> writes:
> On 2020-Feb-24, Tom Lane wrote:
>> Why not just drive it off max_files_per_process?  On Unix, that
>> largely exists to override the ulimit setting anyway.  With no
>> comparable knob on a Windows system, we might as well just say
>> that's what you set.

> That makes sense to me -- but if we do that, then maybe we should be
> doing the setrlimit() dance on it too, on Linux^W^W where supported.

Yeah, arguably we could try to setrlimit if max_files_per_process is
larger than the ulimit.  We should definitely not reduce the ulimit
if max_files_per_process is smaller, though, since the DBA might
intentionally be leaving daylight for purposes such as FD-hungry PL
functions.  On the whole I'm inclined to leave well enough alone on
the Unix side --- there's nothing there that the DBA can't set if
she wishes.

			regards, tom lane