From: Jeff Moyer <jmoyer@redhat.com>
To: Corrado Zoccolo <czoccolo@gmail.com>
Cc: Linux-Kernel <linux-kernel@vger.kernel.org>,
       Jens Axboe <jens.axboe@oracle.com>, Vivek Goyal <vgoyal@redhat.com>
Subject: Re: [PATCH] cfq-iosched: reduce write depth only if sync was delayed
References: <4b1a3d84.0d0db80a.39d0.ffff814a@mx.google.com>
	<x49hbs1fir5.fsf@segfault.boston.devel.redhat.com>
	<4e5e476b0912081246w32a9ae1bw7d49b0285719b22e@mail.gmail.com>
Date: Wed, 09 Dec 2009 13:16:50 -0500
In-Reply-To: <4e5e476b0912081246w32a9ae1bw7d49b0285719b22e@mail.gmail.com>
	(Corrado Zoccolo's message of "Tue, 8 Dec 2009 21:46:39 +0100")
Message-ID: <x49k4wwnhbh.fsf@segfault.boston.devel.redhat.com>
User-Agent: Gnus/5.110011 (No Gnus v0.11) Emacs/23.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1260
Lines: 33

Corrado Zoccolo <czoccolo@gmail.com> writes:

> Hi Jeff,
> On Tue, Dec 8, 2009 at 7:00 PM, Jeff Moyer <jmoyer@redhat.com> wrote:
>> Corrado Zoccolo <czoccolo@gmail.com> writes:
>>
>>> The introduction of ramp-up formula for async queue depths has
>>> slowed down dirty page reclaim, by reducing async write performance.
>>> This patch makes sure the formula kicks in only when sync request
>>> was recently delayed.
>>> @@ -3706,7 +3707,7 @@ static void *cfq_init_queue(struct request_queue *q)
>>>       cfqd->cfq_latency = 1;
>>>       cfqd->cfq_group_isolation = 0;
>>>       cfqd->hw_tag = -1;
>>> -     cfqd->last_end_sync_rq = jiffies;
>>> +     cfqd->last_delayed_sync = jiffies - HZ;
>>
>> So, umm, what's that about?
>
> Previously, when cfq started, it started in a state where writes were
> completely throttled. Now, we optimistically prefer to start with a
> reasonable max depth (10)

OK.  Can we put a comment in there and change the initialization to
cfq_slice_sync * 10?

Thanks,
Jeff
--
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/