From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:54669) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VkyPQ-0003AI-Cx for qemu-devel@nongnu.org; Mon, 25 Nov 2013 10:49:18 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1VkyPK-0006rS-ER for qemu-devel@nongnu.org; Mon, 25 Nov 2013 10:49:12 -0500 Received: from mx1.redhat.com ([209.132.183.28]:42729) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VkyPK-0006rE-4A for qemu-devel@nongnu.org; Mon, 25 Nov 2013 10:49:06 -0500 Message-ID: <52937167.2050508@redhat.com> Date: Mon, 25 Nov 2013 16:48:55 +0100 From: Paolo Bonzini MIME-Version: 1.0 References: <1385387840-17307-1-git-send-email-pl@kamp.de> <1385387840-17307-4-git-send-email-pl@kamp.de> <529364B5.6000202@redhat.com> <529367C8.7080301@kamp.de> <52936951.3090602@redhat.com> <52936BB0.1020608@kamp.de> In-Reply-To: <52936BB0.1020608@kamp.de> Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 1.8 3/6] qemu-img: add option to specify alternate iobuffer size List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Lieven Cc: kwolf@redhat.com, qemu-devel@nongnu.org, stefanha@redhat.com Il 25/11/2013 16:24, Peter Lieven ha scritto: > On 25.11.2013 16:14, Paolo Bonzini wrote: >> Il 25/11/2013 16:07, Peter Lieven ha scritto: >>>>> since the convert process is basically a sync operation it might >>>>> be benificial in some case to change the hardcoded I/O buffer >>>>> size to an alternate (greater) value. >>>> Do you really need the extra knob? You can just add to BlockLimits the >>>> optimal transfer length, and use it unconditionally. >>> If you say patch 5 and 3 are ok. What could be done is to remove >>> this knob and increase the iobuf_size to cluster_size if cluster_size >>> is greater. >> Yes, that makes sense because it avoids unnecessary COW. >> >>> I do not want to increase the default iobuf size to anything >>> greater than 2MB. I do not know why this was choosen, but maybe >>> there was a reason for it. >> I think it is fine to increase it to the cluster_size or even to the >> optimal transfer length (new BlockLimits field). > okay scsi speaking: > bs->bl.optimal_transfer_length = max(iscsilun->bl.opt_xfer_len, > iscsilun->bl.opt_unmap_gran) ? I was thinking more of bs->bl.optimal_transfer_length = lun2qemu(iscsilun->bl.opt_xfer_len); ... iobuf_size = max(bs->bl.optimal_transfer_length, cluster_sectors)*512; iobuf_size = min(16MB, max(2MB, iobuf_size)); > bdi->cluster_size as in Patch 3? Yes, that one's fine. >> * clamp maximum size to optimal transfer length. > or increase it if its larger. for the dell equallogic storages we use > the opt_unmap_gran > is 15MB!!! You're right, maximum size is really bounded anyway by iobuf_size. >> * then, round final sector down to unmap granularity > okay, i will try your modification to only round down the last sector. > would you use bdi->cluster_size or the unmap alignment field from the > bs->bl? bdi->cluster_size. The qemu-img reason for the rounding is to avoid unnecessary COW, which is what bdi->cluster_size is about. It just so happens that it helps your benchmark too. :) Paolo > Peter