Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id <S266689AbRGKOCZ>; Wed, 11 Jul 2001 10:02:25 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org
	id <S266697AbRGKOCF>; Wed, 11 Jul 2001 10:02:05 -0400
Received: from ns.virtualhost.dk ([195.184.98.160]:3848 "EHLO virtualhost.dk")
	by vger.kernel.org with ESMTP id <S266689AbRGKOCC>;
	Wed, 11 Jul 2001 10:02:02 -0400
Date: Wed, 11 Jul 2001 16:01:48 +0200
From: Jens Axboe <axboe@suse.de>
To: Dipankar Sarma <dipankar@sequent.com>
Cc: Mike Anderson <mike.anderson@us.ibm.com>, linux-kernel@vger.kernel.org
Subject: Re: io_request_lock patch?
Message-ID: <20010711160148.A712@suse.de>
In-Reply-To: <20010710172545.A8185@in.ibm.com> <20010710160512.A25632@us.ibm.com> <20010711142311.B9220@in.ibm.com> <20010711105339.F17314@suse.de> <20010711193256.G9220@in.ibm.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20010711193256.G9220@in.ibm.com>
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org

On Wed, Jul 11 2001, Dipankar Sarma wrote:
> On Wed, Jul 11, 2001 at 10:53:39AM +0200, Jens Axboe wrote:
> > The queue lengths should always be long enough to keep the hw busy of
> > course. And in addition, the bigger the queues the bigger the chance of
> > skipping seeks due to reordering. But don't worry, I've scaled the queue
> > lengths so I'm pretty sure that they are always on the safe side in
> > size.
> > 
> > It's pretty easy to test for yourself if you want, just change
> > QUEUE_NR_REQUESTS in blkdev.h. It's currently 8192, the request slots
> > are scaled down from this value. 8k will give you twice the amount of
> > slots that you have RAM in mb, ie 2048 on a 1gig machine.
> > 
> > block: queued sectors max/low 683554kB/552482kB, 2048 slots per queue
> 
> Hmm.. The tiobench run was done on a 1GB machine and we still ran
> out of request slots. Will investigate.

Sure, that's to be expected. If we never ran out we would be wasting
memory. My point is that you should rerun the same test with more
request slots -- and I'd be surprised if you gained any significant
performance on that account. I never said that you'd never run out,
that's of course not true. In fact, running out is what starts the I/O
typically on a 1GB machine and bigger.

-- 
Jens Axboe

-
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/