From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268258AbUH3TSA (ORCPT ); Mon, 30 Aug 2004 15:18:00 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268275AbUH3TR7 (ORCPT ); Mon, 30 Aug 2004 15:17:59 -0400 Received: from omx1-ext.sgi.com ([192.48.179.11]:20895 "EHLO omx1.americas.sgi.com") by vger.kernel.org with ESMTP id S268258AbUH3TR5 (ORCPT ); Mon, 30 Aug 2004 15:17:57 -0400 Date: Mon, 30 Aug 2004 14:10:32 -0500 From: John Hesterberg To: Andrew Morton Cc: Peter Williams , jlan@engr.sgi.com, linux-kernel@vger.kernel.org, erikj@dbear.engr.sgi.com, limin@engr.sgi.com, lse-tech@lists.sourceforge.net, Tim Schmielau Subject: Re: [Lse-tech] Re: [PATCH] new CSA patchset for 2.6.8 Message-ID: <20040830191032.GA5255@sgi.com> References: <412D2E10.8010406@engr.sgi.com> <20040825221842.72dd83a4.akpm@osdl.org> <20040826183834.GA11393@sgi.com> <412EADBC.60607@bigpond.net.au> <20040826205349.0582d38e.akpm@osdl.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040826205349.0582d38e.akpm@osdl.org> User-Agent: Mutt/1.4.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Aug 26, 2004 at 08:53:49PM -0700, Andrew Morton wrote: > Thanks, guys. So we now know that there are three potential > implementations which do much the same thing, yes? I believe CSA does than the others. > I didn't get a sense of a preferred direction, but at least nobody is > flaming anybody else yet ;) > > It strikes me that CSA is the most actively developed and is the furthest > along. But that enhancing BSD accounting might be the least intrusive and > most back-compatible approach. > > Is that a fair summary? If not, what should I have said? Does anyone know if CSA is a super-set of BSD accounting and ELSA? What would be missing? I'm unconvinced that enhancing BSD accounting to encompass the capabilities of CSA is appropriate. I think we can make the data collection additions common. That should encompass the bulk of the invasive changes that are required by at least CSA proper (ie there are still the PAGG changes for job support that we can discuss separately). Not sure about BSD accounting and ELSA. With that cooperation, we can then either proceed with further cooperation, or if the goals and users of the different accounting approaches dictate different kernel modules and user support, I'd propose that might be OK. John