From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from localhost (elowe@localhost) by myrile.madriver.k12.oh.us (8.9.3/8.9.3) with ESMTP id HAA39040 for ; Wed, 11 Oct 2000 07:34:08 -0400 (EDT) (envelope-from elowe@myrile.madriver.k12.oh.us) Date: Wed, 11 Oct 2000 07:34:08 -0400 (EDT) From: Eric Lowe Subject: page-cluster tuning Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org Return-Path: To: linux-mm@kvack.org List-ID: Hi, Over the weekend I played with page-cluster after booting my box with mem=8M. I ran a kernel build and updatedb simultaneously driving it into the swap rather heavily during I/O and found that _both_ processes made much more progress with page-cluster set to 8 than to 4, and the default of 2 was painfully slow because it didn't swap agressively enough. I have yet to do any streaming I/O while swapping tests, but should get to it later in the week. Would anybody like to confirm my results that 8 appears to be an optimum value for page-cluster in 8MB? More to come.. -- Eric Lowe Software Engineer, Systran Corporation elowe@systran.com -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux.eu.org/Linux-MM/