From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753282AbXLAJvX (ORCPT ); Sat, 1 Dec 2007 04:51:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752205AbXLAJvP (ORCPT ); Sat, 1 Dec 2007 04:51:15 -0500 Received: from e33.co.us.ibm.com ([32.97.110.151]:43299 "EHLO e33.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752161AbXLAJvO (ORCPT ); Sat, 1 Dec 2007 04:51:14 -0500 Message-ID: <47512E8D.6090805@linux.vnet.ibm.com> Date: Sat, 01 Dec 2007 15:21:09 +0530 From: Balbir Singh Reply-To: balbir@linux.vnet.ibm.com Organization: IBM User-Agent: Thunderbird 2.0.0.6 (X11/20071022) MIME-Version: 1.0 To: Paul Menage CC: vatsa@linux.vnet.ibm.com, Andrew Morton , dhaval@linux.vnet.ibm.com, containers@lists.linux-foundation.org, linux-kernel@vger.kernel.org, skumar@linux.vnet.ibm.com, torvalds@linux-foundation.org, mingo@elte.hu Subject: Re: [PATCH] sched: cpu accounting controller (V2) References: <6599ad830711122125u576e85a6w428466a0ab46dbc6@mail.gmail.com> <20071113060038.GC3359@linux.vnet.ibm.com> <6599ad830711122205g88aae4fua8dd76cf6e8ab84d@mail.gmail.com> <20071113074805.GA13499@linux.vnet.ibm.com> <6599ad830711122357i60482475o10c0e0935a9e00c0@mail.gmail.com> <20071129191737.GH5681@linux.vnet.ibm.com> <20071129113035.bbdf35db.akpm@linux-foundation.org> <20071129201833.GA18023@linux.vnet.ibm.com> <20071130124225.GM5681@linux.vnet.ibm.com> <6599ad830711302348n49900ca8jfa66fa71fde76a67@mail.gmail.com> In-Reply-To: <6599ad830711302348n49900ca8jfa66fa71fde76a67@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Paul Menage wrote: > Hi Vatsa, > > Thanks, this looks pretty good. > > On Nov 30, 2007 4:42 AM, Srivatsa Vaddagiri wrote: >> - Removed load average information. I felt it needs more thought (esp >> to deal with SMP and virtualized platforms) and can be added for >> 2.6.25 after more discussions. > > The "load" value was never a load average, it was just a count of the > % cpu time used in the previous fixed window of time, updated at the > end of each window. > > Maybe we can instead do something based tracking the length of the run > queue for the cgroup? > > Paul Length of the runqueue gives no idea of the weight of the tasks on the runqueue. I still prefer to have a top'ish view of the load on the system. The load average can be extracted using container group statistics. -- Warm Regards, Balbir Singh Linux Technology Center IBM, ISTL