Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 16 Oct 2002 16:54:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 16 Oct 2002 16:54:38 -0400 Received: from tmr-02.dsl.thebiz.net ([216.238.38.204]:6149 "EHLO gatekeeper.tmr.com") by vger.kernel.org with ESMTP id ; Wed, 16 Oct 2002 16:54:37 -0400 Date: Wed, 16 Oct 2002 17:00:12 -0400 (EDT) From: Bill Davidsen To: "Joseph D. Wagner" cc: "'David S. Miller'" , robm@fastmail.fm, hahn@physics.mcmaster.ca, linux-kernel@vger.kernel.org, jhoward@fastmail.fm Subject: RE: Strange load spikes on 2.4.19 kernel In-Reply-To: <000c01c2728d$263c0ca0$7443f4d1@joe> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1268 Lines: 28 On Sun, 13 Oct 2002, Joseph D. Wagner wrote: > Now wait a minute! Allocating blocks and inodes is an integral part of > a write. Oh sure the actual writing of file data is SMP, but that > process is bottlenecked by single threaded allocation of blocks and > inodes. Perhaps I could phrase what I said to be more technically > accurate by saying, "Writing makes such poor use of multi-threading on > SMP that in terms of performance it's as if it was single threaded." People should note that the reason this hasn't been addressed to date is that the disk is so many orders of magnitude slower than CPU that the practical effect of the "bottleneck" is below the noise with most hardware. So you're not wrong, but you seem to be making it more of an issue than the actual impact seems to justify. I bet you can measure it, or even config a system where it matters, but in most cases it doesn't. -- bill davidsen CTO, TMR Associates, Inc Doing interesting things with little computers since 1979. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/