From: Andres Freund <andres@anarazel.de>
To: Tom Lane <tgl@sss.pgh.pa.us>
Subject: Re: Defaulting wal_sync_method to fdatasync on Linux for 9.1?
Date: Wed, 17 Nov 2010 02:12:29 +0100
User-Agent: KMail/1.13.5 (Linux/2.6.37-rc1-andres-00542-g0143832; KDE/4.5.1;
	x86_64; ; )
Cc: pgsql-performance@postgresql.org, Josh Berkus <josh@agliodbs.com>,
	Greg Smith <greg@2ndquadrant.com>
References: <AANLkTi=n3zxiLNFhbpiN=JPQPwB=BhXf9iepw1KOEs8U@mail.gmail.com>
	<201011170201.24782.andres@anarazel.de>
	<11113.1289955868@sss.pgh.pa.us>
In-Reply-To: <11113.1289955868@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: Text/Plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Message-Id: <201011170212.30033.andres@anarazel.de>

On Wednesday 17 November 2010 02:04:28 Tom Lane wrote:
> Andres Freund <andres@anarazel.de> writes:
> > On Wednesday 17 November 2010 01:51:28 Tom Lane wrote:
> >> Well, there's a forced fsync after writing the last page of an xlog
> >> file, but I don't believe that proves that more than 16MB of xlog
> >> buffers is useless.  Other processes could still be busy filling the
> >> buffers.
> > 
> > Maybe I am missing something, but I think the relevant
> > AdvanceXLInsertBuffer() is currently called with WALInsertLock held?
> 
> The fsync is associated with the write, which is not done with insert
> lock held.  We're not quite that dumb.
Ah, I see. The XLogWrite in AdvanceXLInsertBuffer is only happening if the head 
of the buffer gets to the tail - which is more likely if the wal buffers are 
small...

Andres