From mboxrd@z Thu Jan 1 00:00:00 1970 From: Juan Quintela Subject: KVM call minutes June 29 Date: Tue, 03 Jul 2012 12:18:45 +0200 Message-ID: <87fw99uoze.fsf@elfo.mitica> Reply-To: quintela@redhat.com Mime-Version: 1.0 Content-Type: text/plain To: KVM devel mailing list , qemu-devel@nongnu.org Return-path: Received: from mx1.redhat.com ([209.132.183.28]:49756 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755893Ab2GCKTM (ORCPT ); Tue, 3 Jul 2012 06:19:12 -0400 Sender: kvm-owner@vger.kernel.org List-ID: Isaku reminds me that I forgot to pust that minutes: 2012-06-19 ---------- - migration * xbrle: ok * huge memory: needs migration-thread, but should be ok * postcopy: ask for latency: maximum/avg/std deviation ask for vcpu utilization RDMA? * xbrle don't fix all the problems (vinod) * Freeze is August 1st * They asked for better numbers, at least: * vcpu utilization (in percentange, Avi) * latency of page faults over the network (Anthony) * I can integrate your refactorings (postfix) for 1.2 (I have no problems with them, 19 1st patches or so, but that is up-to-you and the others). - multithreading vhost * vhost has a lot of challenges with lots of small packets. * use a couple of threads by device, one for reception and another for sending (Anthony) * two problems to address (NUMA locality), and scalability (mst) (Anthony has better notes of the last part). Note that I am holidays this week (July 2th), so expect lags on my email response. Later, Juan.