From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:58478) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1d3Paa-0001Xd-2M for qemu-devel@nongnu.org; Wed, 26 Apr 2017 12:14:48 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1d3PaW-00018B-SQ for qemu-devel@nongnu.org; Wed, 26 Apr 2017 12:14:48 -0400 Received: from mx1.redhat.com ([209.132.183.28]:47405) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1d3PaW-000183-MG for qemu-devel@nongnu.org; Wed, 26 Apr 2017 12:14:44 -0400 Date: Wed, 26 Apr 2017 17:14:40 +0100 From: "Dr. David Alan Gilbert" Message-ID: <20170426161439.GG2098@work-vm> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] VM-Migration List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Pradeep Kiruvale Cc: "qemu-devel@nongnu.org Developers" , quintela@redhat.com * Pradeep Kiruvale (pradeepkiruvale@gmail.com) wrote: > Hi All, > > Is it possible to enable/disable to migration features on the fly for a > qemu VM? > I mean through qmp interface kind of a interface? > > I am asking this because keeping the dirty memory list, this could consume > some > compute and memory. So, the plan is to enable when the VM needs to be > migrated. > > Please clarify. These bitmaps should already only be allocated during the migration process, so you shouldn't be seeing any overhead during normal running. See migration/ram.c ram_save_setup->ram_state_init which allocates ram_bitmap and that only happens at the start of migration. ram_state_init also calls memory_global_dirty_log_start which does the kernel side of it. Dave > Regards, > Pradeep -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK