From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35849) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cr4NS-0005aa-LT for qemu-devel@nongnu.org; Thu, 23 Mar 2017 11:10:20 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cr4NQ-0002ed-BD for qemu-devel@nongnu.org; Thu, 23 Mar 2017 11:10:14 -0400 Date: Thu, 23 Mar 2017 15:09:55 +0000 From: Stefan Hajnoczi Message-ID: <20170323150955.GD12560@stefanha-x1.localdomain> References: <1489233245-27057-1-git-send-email-ashijeetacharya@gmail.com> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="F8dlzb82+Fcn6AgP" Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] [PATCH 0/2] Allocate mutiple clusters for VMDK I/O List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Ashijeet Acharya Cc: Fam Zheng , John Snow , Kevin Wolf , Max Reitz , qemu block , qemu-devel --F8dlzb82+Fcn6AgP Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, Mar 21, 2017 at 09:14:08AM +0000, Ashijeet Acharya wrote: > On Tue, 21 Mar 2017 at 13:21, Stefan Hajnoczi wrote: >=20 > > On Sat, Mar 11, 2017 at 11:54 AM, Ashijeet Acharya > > wrote: > > > This series optimizes the I/O performance of VMDK driver. > > > > > > Patch 1 makes the VMDK driver to allocate multiple clusters at once. > > Earlier > > > it used to allocate cluster by cluster which slowed down its performa= nce > > to a > > > great extent. > > > > > > Patch 2 changes the metadata update code to update the L2 tables for > > multiple > > > clusters at once. > > > > This patch series is a performance optimization. Benchmark results > > are required to justify optimizations. Please include performance > > results in the next revision. > > > > A popular disk I/O benchmarking is fio (https://github.com/axboe/fio). > > I suggest a write-heavy workload with a large block size: > > > > $ cat fio.job > > [global] > > direct=3D1 > > filename=3D/dev/vdb > > ioengine=3Dlibaio > > runtime=3D30 > > ramp_time=3D5 > > > > [job1] > > iodepth=3D4 > > rw=3Drandwrite > > bs=3D256k > > $ for i in 1 2 3 4 5; do fio --output=3Dfio-$i.txt fio.job; done # > > WARNING: overwrites /dev/vdb > > > > It's good practice to run the benchmark several times because there is > > usually some variation between runs. This allows you to check that > > the variance is within a reasonable range (5-10% on a normal machine > > that hasn't been specially prepared for benchmarking). >=20 >=20 > I ran a few write tests of 128M using qemu-io and the results showed the > time to drop to almost half, will those work? Although, I will also try to > use the tool you mentioned later today when I am free and include those > results as well. Maybe, it's hard to say without seeing the commands you ran. Stefan --F8dlzb82+Fcn6AgP Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- iQEcBAEBAgAGBQJY0+VDAAoJEJykq7OBq3PI2EMH/3ADbT/W4mGhFZ9DIi95Lr3q 2KEB8qBf0Didrx+rg7cGpPlXfKqgEagqFBaQbzepILkw25QUUPT6vgqGefYEHdeT 3H3O32vvwcasEbMCanH6exqPr9x5W3cjrg0QlfcBBRlx9inWo5rylaz63m7pS8Wk V5b3y8lfhK7/AvMOpUraCDZEbnTWXytFBFJ3cOK6E/zd54wm/51dERKNwlXoLuNO 9dXP9gIMSJEIE86Fo0C5V30luGxtpFw/L3FucuGKcJN3ognQz3CZAwTKyP6B3cXi L7M0AIGBoZyBc/Z+AR5+jVVJ+VItOqSmAsVwL/m82nKih9zrHALkOv7mKN1WAYI= =7fQz -----END PGP SIGNATURE----- --F8dlzb82+Fcn6AgP--