From owner-freebsd-questions Thu Jan 12 09:58:00 1995 Return-Path: questions-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id JAA26777 for questions-outgoing; Thu, 12 Jan 1995 09:58:00 -0800 Received: from wavefront.wti.com (WAVEFRONT.WTI.COM [144.253.1.1]) by freefall.cdrom.com (8.6.9/8.6.6) with SMTP id JAA26771 for ; Thu, 12 Jan 1995 09:57:59 -0800 Received: from walrus.wti.com by wavefront.wti.com (4.1/SMI-1.0-WTI Special) id AA00740; Thu, 12 Jan 95 09:57:48 PST Received: by walrus.wti.com (931110.SGI/Wavefront-Client-2.0) id AA02722; Thu, 12 Jan 95 09:57:08 -0800 From: "Steve Galle" Message-Id: <9501120957.ZM2720@walrus.wti.com> Date: Thu, 12 Jan 1995 09:57:08 -0800 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: FreeBSD-Questions@FreeBSD.org Subject: mfs file system speed Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Sender: questions-owner@FreeBSD.org Precedence: bulk Hi all. I just took the time to mount /tmp as an mfs file system, and have made some interesting measurements. It seems that on my machine (a 486DX4/100) writes to a mfs file system are slower than writes to a real file system. I made the changes to my file system after hearing that access to the mfs files should be faster. A 26 meg file took 23 seconds to write to /, and over a minute to write to the mfs partition. Can anyone shed some light on this? Thanks! -Steve Galle steveg@wavefront.wti.com