From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 25 Sep 2001 12:03:12 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 25 Sep 2001 12:03:03 -0400 Received: from prgy-npn1.prodigy.com ([207.115.54.37]:17936 "EHLO deathstar.prodigy.com") by vger.kernel.org with ESMTP id ; Tue, 25 Sep 2001 12:02:50 -0400 Date: Tue, 25 Sep 2001 12:03:12 -0400 Message-Id: <200109251603.f8PG3Cj06561@deathstar.prodigy.com> To: linux-kernel@vger.kernel.org Subject: Re: Linux VM design X-Newsgroups: linux.dev.kernel In-Reply-To: <20010924182948Z16175-2757+1593@humbolt.nl.linux.org> Organization: TMR Associates, Schenectady NY From: davidsen@tmr.com (bill davidsen) Reply-To: davidsen@tmr.com (bill davidsen) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org In article <20010924182948Z16175-2757+1593@humbolt.nl.linux.org> phillips@bonn-fries.net wrote: | You might want to have a look at this: | | http://archi.snu.ac.kr/jhkim/seminar/96-004.ps | (lrfu algorithm) | | To tell the truth, I don't really see why the frequency information is all | that useful either. Rik suggested it's good for streaming IO but we already | have effective means of dealing with that that don't rely on any frequency | information. A count which may actually be useful is a count of how many time the page has been swapped in (after being swapped out) as a predictor that it will be a good page to keep. The problem is that there are many things which help, and I don't think we have the balance quite right yet. I suspect that there need to be some hysteresis and runtime tuning over seconds to get optimal performance. Of course systems with really odd loads will still need to have hand tuning, and the /proc/sys interface should include sensible ways to do this. | So the list of reasons why aging is good is looking really short. The primary reason on my list is that under some load conditions it produces much better response. Note that I didn't say all conditions before you rush to disagree with me. Sometimes people will trade a little steady state performance to avoid a really bad worst case. How the problem is solved really isn't the issue, but responsiveness is important. Right now it seems some people are reporting that their loads work better with aging. -- bill davidsen "If I were a diplomat, in the best case I'd go hungry. In the worst case, people would die." -- Robert Lipe