From owner-freebsd-hackers@FreeBSD.ORG Thu Sep 29 18:58:38 2005 Return-Path: X-Original-To: freebsd-hackers@freebsd.org Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id F07EE16A41F for ; Thu, 29 Sep 2005 18:58:38 +0000 (GMT) (envelope-from keramida@linux.gr) Received: from rosebud.otenet.gr (rosebud.otenet.gr [195.170.0.94]) by mx1.FreeBSD.org (Postfix) with ESMTP id EAE4043D62 for ; Thu, 29 Sep 2005 18:58:34 +0000 (GMT) (envelope-from keramida@linux.gr) Received: from flame.pc (aris.bedc.ondsl.gr [62.103.39.226]) by rosebud.otenet.gr (8.13.4/8.13.4/Debian-1) with SMTP id j8TIwQXD031390; Thu, 29 Sep 2005 21:58:27 +0300 Received: from flame.pc (flame [127.0.0.1]) by flame.pc (8.13.4/8.13.4) with ESMTP id j8TIvo12015755; Thu, 29 Sep 2005 21:57:50 +0300 (EEST) (envelope-from keramida@linux.gr) Received: (from keramida@localhost) by flame.pc (8.13.4/8.13.4/Submit) id j8TIvnRt015754; Thu, 29 Sep 2005 21:57:49 +0300 (EEST) (envelope-from keramida@linux.gr) Date: Thu, 29 Sep 2005 21:57:49 +0300 From: Giorgos Keramidas To: Eric Anderson Message-ID: <20050929185749.GB15714@flame.pc> References: <433C30AB.2030409@centtech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <433C30AB.2030409@centtech.com> Cc: freebsd-hackers@freebsd.org Subject: Re: ufsstat implementation X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 29 Sep 2005 18:58:39 -0000 On 2005-09-29 13:21, Eric Anderson wrote: > I've been looking at the ufs code, and thinking about wedging in some > statistic gathering pieces pretty much copied from the nfs code (and > nfsstat) to provide nearly the same functionality. > > I realize that adding statistic gathering code would techinically reduce > filesystem performance, so should I put a wrapper around the code pieces > checking a sysctl, or use #ifdef's around it, or neither? It would take running an implementation with and without ufsstats to see what the impact on filesystem performance is, so it seems to me the natural choise for early stages of development is an #ifdef that can easily be turned on/off and control where all of the code goes in or nothing at all :-)