Date: Fri, 11 Jun 2021 19:43:47 -0500
From: Justin Pryzby <pryzby@telsasoft.com>
To: Euler Taveira <euler.taveira@2ndquadrant.com>
Cc: Tomas Vondra <tomas.vondra@2ndquadrant.com>,
	pgsql-hackers@lists.postgresql.org
Subject: Re: cleanup temporary files after crash
Message-ID: <20210612004347.GP16435@telsasoft.com>
References: 
 <CAH503wDKdYzyq7U-QJqGn=Gm6XmoK+6_6xTJ-Yn5WSvoHLY1Ww@mail.gmail.com>
 <20201028184252.5vpoe65kcfebbmmf@development>
 <CAH503wCg+yne7hKAAFFHsVPe+tPb-aF5VGaWG0ETqyB3GP=Q+g@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAH503wCg+yne7hKAAFFHsVPe+tPb-aF5VGaWG0ETqyB3GP=Q+g@mail.gmail.com>
User-Agent: Mutt/1.9.4 (2018-02-28)
Archived-At: 
 <https://www.postgresql.org/message-id/20210612004347.GP16435%40telsasoft.com>
Precedence: bulk

On Sat, Oct 31, 2020 at 09:01:15PM -0300, Euler Taveira wrote:
> > > The current behavior is only useful for debugging purposes.

On Wed, 28 Oct 2020 at 15:42, Tomas Vondra <tomas.vondra@2ndquadrant.com> wrote:
> > One thing I'm not sure about is whether we should have the GUC as
> > proposed, or have a negative "keep_temp_files_after_restart" defaulting
> > to false. But I don't have a very good justification for the alternative
> > other than vague personal preference.

On Sat, Oct 31, 2020 at 09:01:15PM -0300, Euler Taveira wrote:
> I thought about not providing a GUC at all or provide it in the developer
> section. I've never heard someone saying that they use those temporary
> files to investigate an issue. Regarding a crash, all information is already
> available and temporary files don't provide extra details. This new GUC is just to keep the
> previous behavior. I'm fine without the GUC, though.

Should this GUC be classified as a developer option, and removed from
postgresql.sample.conf ?

That was discussed initially in October but not since.

On Wed, Oct 28, 2020 at 11:16:26AM -0300, Euler Taveira wrote:
> It also has an undesirable behavior: you have to restart the service to
> reclaim the space.

BTW, that's not really true - you can remove the tempfiles while the server is
running.  The complainant in bug#16427 was being careful - which is good.
I watch for and remove tempfiles older than 2 days.  The worst consequence of
removing a tempfile would be a failed query.  

-- 
Justin