From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753073AbXCRFzZ (ORCPT ); Sun, 18 Mar 2007 01:55:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753074AbXCRFzZ (ORCPT ); Sun, 18 Mar 2007 01:55:25 -0400 Received: from mtaout2.012.net.il ([84.95.2.4]:59696 "EHLO mtaout2.012.net.il" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753061AbXCRFzY (ORCPT ); Sun, 18 Mar 2007 01:55:24 -0400 Date: Sun, 18 Mar 2007 07:55:21 +0200 From: Avi Kivity Subject: Re: is RSDL an "unfair" scheduler too? In-reply-to: <20070318052439.GT943@1wt.eu> To: Willy Tarreau Cc: Linus Torvalds , William Lee Irwin III , Ingo Molnar , Con Kolivas , ck@vds.kolivas.org, Serge Belyshev , Al Boldi , Mike Galbraith , linux-kernel@vger.kernel.org, Nicholas Miell , Andrew Morton Message-id: <45FCD449.60303@argo.co.il> MIME-version: 1.0 Content-type: text/plain; charset=ISO-8859-1; format=flowed Content-transfer-encoding: 7BIT X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-2.1.12 (firebolt.argo.co.il [0.0.0.0]); Sun, 18 Mar 2007 07:55:22 +0200 (IST) References: <200703042335.26785.a1426z@gawab.com> <20070317074506.GA13685@elte.hu> <87fy84i7nn.fsf@depni.sinp.msu.ru> <200703172048.46267.kernel@kolivas.org> <20070317114903.GA20673@elte.hu> <45FC525D.5000708@argo.co.il> <20070318012533.GB2986@holomorphy.com> <20070318052439.GT943@1wt.eu> User-Agent: Thunderbird 1.5.0.9 (X11/20070212) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Willy Tarreau wrote: > On Sat, Mar 17, 2007 at 06:32:29PM -0700, Linus Torvalds wrote: > >> On Sat, 17 Mar 2007, William Lee Irwin III wrote: >> >>> One issue this raises is prioritizing users on a system, threads within >>> processes, jobs within users, etc. >>> >> Doing some "classing" even by just euid might be a good idea. It would >> actually catch X automatically most of the time, because the euid of the X >> server is likely to be root, so even for the "trivial" desktop example, it >> would kind of automatically mean that X would get about 50% of CPU time >> even if you have a hundred user clients, just because that's "fair" by >> euid. >> > > Warning: all these ideas seem interesting for desktop, but are definitely > not for servers. I found RSDL to be excellent on servers, compared to > mainline in which some services are starving under load. I can understand > that on the desktop people want some unfairness, and I like the pgrp idea > for instance. But this one will certainly fail on servers, or make the > admins get grey hair very soon. > I didn't suggest adding any unfairness! I suggested being fair by user/job/process instead of being fair by thread (which is actually unfair as it favors multi threaded processes over single threaded processes). > Maybe we're all discussing the problem because we have reached the point > where we need two types of schedulers : one for the desktop and one for > the servers. After all, this is already what is proposed with preempt, > it would make sense provided they share the same core and avoid ifdefs > or unused structure members. Maybe adding OPTIONAL unfairness to RSDL > would help some scenarios, but in any case it is important to retain > the default fairness it provides. > I hope not. I think that reducing the timeslice base, combined with renicing X all the way to hell should suffice. -- Do not meddle in the internals of kernels, for they are subtle and quick to panic.