From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail137.messagelabs.com (mail137.messagelabs.com [216.82.249.19]) by kanga.kvack.org (Postfix) with ESMTP id A2ECC6B002D for ; Sat, 22 Oct 2011 05:26:33 -0400 (EDT) Date: Sat, 22 Oct 2011 11:26:16 +0200 From: Michal Hocko Subject: Re: [RFD] Isolated memory cgroups again Message-ID: <20111022092524.GA5497@tiehlicka.suse.cz> References: <20111020013305.GD21703@tiehlicka.suse.cz> <20111020105950.fd04f58f.kamezawa.hiroyu@jp.fujitsu.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: owner-linux-mm@kvack.org List-ID: To: Balbir Singh Cc: KAMEZAWA Hiroyuki , linux-mm@kvack.org, LKML , Johannes Weiner , Daisuke Nishimura , Hugh Dickins , Ying Han , Andrew Morton , Glauber Costa , Kir Kolyshkin , Pavel Emelianov , GregThelen , "pjt@google.com" , Tim Hockin , Dave Hansen , Paul Menage , James Bottomley On Fri 21-10-11 21:34:06, Balbir Singh wrote: > On Thu, Oct 20, 2011 at 7:29 AM, KAMEZAWA Hiroyuki > wrote: > > On Wed, 19 Oct 2011 18:33:09 -0700 > > Michal Hocko wrote: > > > >> Hi all, > >> this is a request for discussion (I hope we can touch this during memcg > >> meeting during the upcoming KS). I have brought this up earlier this > >> year before LSF (http://thread.gmane.org/gmane.linux.kernel.mm/60464). > >> The patch got much smaller since then due to excellent Johannes' memcg > >> naturalization work (http://thread.gmane.org/gmane.linux.kernel.mm/68724) > >> which this is based on. > > > > Hi, Michal Hi Balbir, > > I'd like to understand, what the isolation is for? > > 1. Is it an alternative to memory guarantees? Not really, it is more about resident working set guarantee and workload isolations wrt. memory. > 2. How is this different from doing cpusets (fake NUMA) and isolating them? Yes this would work. I have not many experiences in this area but I guess the primary stopper for fake NUMA is that it is x86_64 only, configuration is static and little bit awkward to use (nodes of the same size e.g.). I understood that google is moving out of fake NUMA towards memcg for those reasons. > > Just trying to catch up, > Balbir -- Michal Hocko SUSE Labs SUSE LINUX s.r.o. Lihovarska 1060/12 190 00 Praha 9 Czech Republic -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: email@kvack.org