From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261944AbUL0Sjx (ORCPT ); Mon, 27 Dec 2004 13:39:53 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261946AbUL0Sjx (ORCPT ); Mon, 27 Dec 2004 13:39:53 -0500 Received: from smtprelay04.ispgateway.de ([80.67.18.16]:16588 "EHLO smtprelay04.ispgateway.de") by vger.kernel.org with ESMTP id S261944AbUL0Sjv convert rfc822-to-8bit (ORCPT ); Mon, 27 Dec 2004 13:39:51 -0500 From: Ingo Oeser To: linux-kernel@vger.kernel.org Subject: Re: cache and buffer usage with kernel 2.6.10 Date: Mon, 27 Dec 2004 19:08:29 +0100 User-Agent: KMail/1.6.2 References: In-Reply-To: MIME-Version: 1.0 Content-Disposition: inline Content-Type: Text/Plain; charset="big5" Content-Transfer-Encoding: 8BIT Message-Id: <200412271908.31269.ioe-lkml@rameria.de> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Monday 27 December 2004 17:33, M Benz wrote: > Is there any change with memory management btw the two kernels? Also, what > is the different btw "cache" and "buffer"? Why don't kernel 2.6.10 use all > the memory but leave around 200MB free? These metrics have been useless from the very beginning and are not very descriptive. If you like to get some meaning in your statistice, use vmstat. This is more costly, but actually shows something about the performance of your system. Maybe we could concentrate on exporting these metrics instead buffer/cache and provide dummy metrics for them. At least "buffer" can be made to sth. like "shared" does now. Regards Ingo Oeser -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) iD8DBQFB0E+dU56oYWuOrkARAssXAKDEPxGEE+r162ZGRcYTarl7qofvGgCfX2Yt 6IdeNnYV+lZvbgT0trg9xtc= =+6ii -----END PGP SIGNATURE-----