From: Tom Lane <tgl@sss.pgh.pa.us>
To: Christian Ullrich <chris@chrullrich.net>
cc: pgsql-admin@postgresql.org
Subject: Re: Why does the WAL writer sit on completed segments (on Windows)?
In-reply-to: <b53e3f50-54b4-49ad-9344-97494ab5cbbd@chrullrich.net>
References: <od2lad$df0$1@blaine.gmane.org> <17213.1492442911@sss.pgh.pa.us>
 <b53e3f50-54b4-49ad-9344-97494ab5cbbd@chrullrich.net>
Comments: In-reply-to Christian Ullrich <chris@chrullrich.net>
	message dated "Mon, 17 Apr 2017 17:44:53 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <28441.1492445204.1@sss.pgh.pa.us>
Date: Mon, 17 Apr 2017 12:06:44 -0400
Message-ID: <28442.1492445204@sss.pgh.pa.us>
Precedence: bulk
Sender: pgsql-admin-owner@postgresql.org

Christian Ullrich <chris@chrullrich.net> writes:
> So, just to be sure, you're saying that when the ".ready" file is 
> created, it is guaranteed that no process will write to a location 
> within that segment anymore even if it happens to have the file open?
> (That's actually completely obvious, but bear with me, please.)

Right.

> And further, that each such process will close the open file as soon as 
> it, intending to write WAL, finds out that the next available location 
> is not within the file?

Yes, I believe it will typically hold the file open until that happens.
(If short of file descriptors, it might close it sooner, but that's
probably not common.)

			regards, tom lane


-- 
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin