From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265134AbUGZKYz (ORCPT ); Mon, 26 Jul 2004 06:24:55 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265148AbUGZKYz (ORCPT ); Mon, 26 Jul 2004 06:24:55 -0400 Received: from grendel.digitalservice.pl ([217.67.200.140]:38847 "HELO mail.digitalservice.pl") by vger.kernel.org with SMTP id S265134AbUGZKYw (ORCPT ); Mon, 26 Jul 2004 06:24:52 -0400 From: "R. J. Wysocki" Organization: SiSK To: Con Kolivas Subject: Re: Autotune swappiness01 Date: Mon, 26 Jul 2004 12:34:29 +0200 User-Agent: KMail/1.5 Cc: Andrew Morton , linux-kernel@vger.kernel.org References: <200407261052.50178.rjwysocki@sisk.pl> <4104CF8F.2050208@kolivas.org> In-Reply-To: <4104CF8F.2050208@kolivas.org> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-2" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200407261234.29565.rjwysocki@sisk.pl> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Monday 26 of July 2004 11:31, Con Kolivas wrote: > R. J. Wysocki wrote: > > On Monday 26 of July 2004 03:09, Con Kolivas wrote: > >>Con Kolivas writes: > >>>Andrew Morton writes: > >>>>Seriously, we've seen placebo effects before... > >>> > >>>I am in full agreement there... It's easy to see that applications do > >>> not swap out overnight; but i'm having difficulty trying to find a way > >>> to demonstrate the other part. I guess timing the "linking the kernel > >>> with full debug" on a low memory box is measurable. > >> > >>I should have said - finding a swappiness that ensures not swapping out > >>applications with updatedb, then using that same swappiness value to do > >> the linking test. > > > > Please excuse me, but is that viable at all? IMHO, it's just like trying > > to tune a radio including volume with only one knob. I don't say it > > won't work, but the probability that it will is rather small, it seems > > ... > > Well that's what we want. I cant remember other desktop operating > systems setting a root only control between night and day, or between > copying ISOs and running applications or... I agree, but isn't it related to the fact that other desktop OSes usually don't run anything like updatedb nightly? Perhaps we need a bit more sophisticated swap algorithm than other OSes do. For example, couldn't we add an additional parameter to control the swapping "behavior", apart from the swappiness? Something like adding the second knob in my radio example? Just thinking, rjw -- Rafael J. Wysocki [tel. (+48) 605 053 693] ---------------------------- For a successful technology, reality must take precedence over public relations, for nature cannot be fooled. -- Richard P. Feynman