Date: Thu, 10 Jul 2008 06:18:22 -0700
From: Andrew Morton <akpm@linux-foundation.org>
To: Martin Sustrik <sustrik@fastmq.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>, Martin Lucina <mato@kotelna.sk>,
       linux-kernel@vger.kernel.org
Subject: Re: Higher than expected disk write(2) latency
Message-Id: <20080710061822.38975133.akpm@linux-foundation.org>
In-Reply-To: <48760BFB.90007@fastmq.com>
References: <20080628121131.GA14181@nodbug.moloch.sk>
	<20080709222701.8eab4924.akpm@linux-foundation.org>
	<4875C45C.2010901@fastmq.com>
	<20080710093109.46db7c40@the-village.bc.nu>
	<48760BFB.90007@fastmq.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1055
Lines: 25

On Thu, 10 Jul 2008 15:17:47 +0200 Martin Sustrik <sustrik@fastmq.com> wrote:

> Hi Alan,
> 
> >> What we see is that AIO performs rather bad while we are still 
> >> enqueueing more writes (it misses right position on the disk and has to 
> >> do superfluous disk revolvings), however, once we stop enqueueing new 
> >> write request, those already in the queue are processed swiftly.
> > 
> > Which disk scheduler are you using - some of the disk schedulers
> > intentionally delay writes to try and get better block merging.
> 
> It's CFQ. Does it delay writes? And if so, what should we use instead?
> 

noop is the simplest scheduler.  deadline is the simplest real scheduler,
and deadline doesn't have any delaying logic.

If CFQ or anticipatory _are_ putting delays into this workload, that'd be
a bug.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/