From owner-freebsd-hackers Mon Nov 20 06:33:49 1995 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id GAA10960 for hackers-outgoing; Mon, 20 Nov 1995 06:33:49 -0800 Received: from terra.Sarnoff.COM (terra.sarnoff.com [130.33.11.203]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id GAA10946 for ; Mon, 20 Nov 1995 06:33:45 -0800 Received: (from rminnich@localhost) by terra.Sarnoff.COM (8.6.12/8.6.12) id JAA03493; Mon, 20 Nov 1995 09:31:18 -0500 Date: Mon, 20 Nov 1995 09:31:17 -0500 (EST) From: "Ron G. Minnich" To: Terry Lambert cc: "Serge A. Babkin" , hackers@freebsd.org Subject: Re: NFS client caching in UNIX In-Reply-To: <199511182045.NAA09519@phaeton.artisoft.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-hackers@freebsd.org Precedence: bulk one note though: every study i've seen of nfs activity (including my own multi-month monitoring of a net i ran) shows 1) most DATA is for read (last time i checked read outnumbered write by over 10 to 1) 2) most of that read data is to read only file systems (/usr/local, etc.) (think of emacs, frame, your favorite cad system, etc.) 3) most PACKETS are lookup/gettatr, again to read only file systems so you can win a lot by simply running CFS (from sun) or other caching file system. I have a caching file system for nfs i wrote a number of years ago (Autocacher) but it needs work. Anybody want to do the work, you can have the code. It consists of a modified AMD. ron Ron Minnich |Like a knife through Daddy's heart: rminnich@sarnoff.com |"Don't make fun of Windows, daddy! It takes care (609)-734-3120 | of all my files and it's reliable and I like it".