From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:36856) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Z1xNf-0008Om-Ct for qemu-devel@nongnu.org; Mon, 08 Jun 2015 09:46:24 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Z1xNc-0000wZ-1r for qemu-devel@nongnu.org; Mon, 08 Jun 2015 09:46:23 -0400 Received: from mail-wi0-x22b.google.com ([2a00:1450:400c:c05::22b]:35939) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Z1xNb-0000w7-O8 for qemu-devel@nongnu.org; Mon, 08 Jun 2015 09:46:19 -0400 Received: by wigg3 with SMTP id g3so53043317wig.1 for ; Mon, 08 Jun 2015 06:46:18 -0700 (PDT) Sender: Paolo Bonzini Message-ID: <55759CA8.4070909@redhat.com> Date: Mon, 08 Jun 2015 15:46:16 +0200 From: Paolo Bonzini MIME-Version: 1.0 References: <1b79c169d7e76e0ae1dfe3f79b8f9cfe@alukardd.org> In-Reply-To: Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] poor virtio-scsi performance List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Andrey Korolyov , Alexey Cc: "qemu-devel@nongnu.org" On 08/06/2015 15:10, Andrey Korolyov wrote: > > You are probably hitting the (more or less fundamental) ceiling which > was the reason for introducing dataplane backend recently, in other > words, it does not matter how fast your backend is, the operation > number will be limited by 50..100kIOPS without dataplane. As far as I > remember the recent development made possible using it with > virtio-scsi as well as with virtio-blk, so you`d possibly want to try > it. virtio-scsi dataplane is not yet 100% thread-safe, but in practice it should work in 2.3+. It's definitely good enough for benchmarking. Paolo