From owner-freebsd-geom@FreeBSD.ORG Thu Jul 26 13:47:15 2012 Return-Path: Delivered-To: freebsd-geom@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id B0BA6106566C for ; Thu, 26 Jul 2012 13:47:15 +0000 (UTC) (envelope-from gcubfg-freebsd-geom@m.gmane.org) Received: from plane.gmane.org (plane.gmane.org [80.91.229.3]) by mx1.freebsd.org (Postfix) with ESMTP id 649968FC0A for ; Thu, 26 Jul 2012 13:47:15 +0000 (UTC) Received: from list by plane.gmane.org with local (Exim 4.69) (envelope-from ) id 1SuOAd-0005pn-BQ for freebsd-geom@freebsd.org; Thu, 26 Jul 2012 15:32:03 +0200 Received: from lara.cc.fer.hr ([161.53.72.113]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 26 Jul 2012 15:32:03 +0200 Received: from ivoras by lara.cc.fer.hr with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 26 Jul 2012 15:32:03 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: freebsd-geom@freebsd.org From: Ivan Voras Date: Thu, 26 Jul 2012 15:31:51 +0200 Lines: 55 Message-ID: References: <501056C4.3080806@FreeBSD.org> <50106E5F.4030402@FreeBSD.org> <501074E4.4040805@FreeBSD.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="------------enigF19539FA2CE033B976BD1189" X-Complaints-To: usenet@dough.gmane.org X-Gmane-NNTP-Posting-Host: lara.cc.fer.hr User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:10.0) Gecko/20120213 Thunderbird/10.0 In-Reply-To: <501074E4.4040805@FreeBSD.org> X-Enigmail-Version: 1.3.5 Cc: freebsd-scsi@freebsd.org, freebsd-hackers@freebsd.org Subject: Re: geom <-> cam disk X-BeenThere: freebsd-geom@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: GEOM-specific discussions and implementations List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 26 Jul 2012 13:47:15 -0000 This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enigF19539FA2CE033B976BD1189 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable On 26/07/2012 00:36, Andriy Gapon wrote: > on 26/07/2012 01:08 Alexander Motin said the following: >> Different controllers have different command queueing limitations. If = you are >> testing with ahci(4) driver and modern disks, then their 32 command sl= ots per >> port can be enough for many workloads to enqueue all commands to the h= ardware >> and leave queue empty as you've described. But if you take harder work= load, or >> controller/ device without command queueing support, extra requests wi= ll be >> accumulated on that bioq and sorted there. >=20 > Alexander, >=20 > thank you for the reply. > Indeed, using 64 parallel dd processes with bs=3D512 I was able to 'kic= k in' the > disksort logic. But I am not sure if the disksort algorithm makes much= > difference in this case given the number of commands that a disk firmwa= re can > internally re-order. (Not mentioning that potentially disksort could s= tarve > some I/O bound processes in favor of others -- but that's a totally dif= ferent > topic). Is the queue drained in-order (i.e. after sorting)? Then it could help as the device queues will be filled with requests whose offsets are closer together (as opposed to totally random), which helps with bursts (e.g. untarring the ports tree). --------------enigF19539FA2CE033B976BD1189 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.19 (FreeBSD) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAlARRscACgkQ/QjVBj3/HSy7SACeJEx+tZfRdo2stam9R4hWNCUL NzYAnR00QKCAvhtJ3FJ4vAZ3TIf0vZOL =twL2 -----END PGP SIGNATURE----- --------------enigF19539FA2CE033B976BD1189--