From: Marc MERLIN Subject: Re: du -s src is a lot slower on SSD than spinning disk in the same laptop Date: Wed, 25 Jul 2012 16:38:59 -0700 Message-ID: <20120725233859.GA6752@merlins.org> References: <20120725154521.GA3398@merlins.org> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: tytso@mit.edu, linux-ext4@vger.kernel.org, axboe@kernel.dk, Milan Broz To: =?utf-8?B?THVrw6HFoQ==?= Czerner Return-path: Received: from magic.merlins.org ([209.81.13.136]:56511 "EHLO mail1.merlins.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752343Ab2GYXjD (ORCPT ); Wed, 25 Jul 2012 19:39:03 -0400 Content-Disposition: inline In-Reply-To: Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, Jul 25, 2012 at 08:40:28PM +0200, Luk=C3=A1=C5=A1 Czerner wrote= : > yesterday Milan Broz pointed me to the other problem of yours where > by reading from dm_crypt target on the SSD was much slower than on > the spinning disk. I am not sure if he already shared some > information he managed to gather, but we saw that reading from SSD > was much slower probably because reads were divided into tons of > small (page size) bios as opposed to bigger chunks on spinning > disk. =20 Yes, indeed. To make things simpler, I'm not using dmcrypt here. > This is probably the same reason here. The reason is most likely > that we handle SSD differently than spinning disk (turn off > elevator, different io scheduler and possibly other things I am not > aware of). Also IIRC bio merging should be less "aggressive" on SSD > than spinning disk (or maybe even turned off?), because SSD should > supposedly handle much more iops than than spinning drive, hence > waiting for a merge might slow things down. However in this case it > seems to have quite opposite effect for some reason. >=20 > You may try to convince kernel to treat your SSD as rotational disk > by: >=20 > echo 1 > /sys/block/sda/queue/rotational > and see if it improves things. Actually I had done that, along with making the readahead 8K change, bu= t that didn't help: gandalfthegreat:/mnt/mnt2# time du -sh src/ 519M src/ real 0m12.176s gandalfthegreat:/mnt/mnt2# df -h . =46ilesystem Size Used Avail Use% Mounted on /dev/sda2 25G 3.0G 21G 13% /mnt/mnt2 gandalfthegreat:/mnt/mnt2# blockdev --report /dev/sda2 RO RA SSZ BSZ StartSec Size Device rw 8192 512 4096 502272 26843283456 /dev/sda2 gandalfthegreat:/mnt/mnt2# cat /sys/block/sda/queue/rotational 1 gandalfthegreat:/mnt/mnt2#=20 > Unfortunately I think that there is not much we can do about this > from the file system level. Someone from block level should > definitely take a look at this issue. Jens ? Ok, I just wanted to rule out that it was not a VFS issue. If you're confident it's block level and basically having storage that = is too fast (and indeed, I just bought one of the fastest SSDs out there) = is actually causing problems that make the entire system much slower as a result, I'm happy to take it up another list. Where would you recommend I go with this? Thanks, Marc --=20 "A mouse is a device used to point at the xterm you want to type in" - = A.S.R. Microsoft is to operating systems .... .... what McDonalds is to gourmet= cooking Home page: http://marc.merlins.org/ =20 -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html