From owner-freebsd-hackers Sun Apr 13 12:24:37 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA19956 for hackers-outgoing; Sun, 13 Apr 1997 12:24:37 -0700 (PDT) Received: from alpha.risc.org (taob@trt-on22-13.netcom.ca [207.181.87.205]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA19949 for ; Sun, 13 Apr 1997 12:24:31 -0700 (PDT) Received: from localhost (taob@localhost) by alpha.risc.org (8.8.4/8.8.4) with SMTP id PAA21565 for ; Sun, 13 Apr 1997 15:24:21 -0400 (EDT) Date: Sun, 13 Apr 1997 15:24:21 -0400 (EDT) From: Brian Tao To: FREEBSD-HACKERS-L Subject: Maximizing NFS server performance Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Given two fast systems (both at least P200's) and a crossover fast Ethernet cable between them, one an NFS client and the other an NFS server, what options do I have to maximize the performance of the NFS server? It will be running FreeBSD with a couple of ultrawide controllers and a handful of disks attached to it. I need to move about 600,000 files totalling 12 gigabytes from one system to another. Since the directory organization is changing, I can't use FTP or tape backup to migrate the data. Using a loopback NFS mount in a P200 running 3.0-970124-SNAP, it takes almost two minutes to copy my 23MB /usr/src/sys hierarchy (2277 files). A dd from /dev/zero to a file on the same NFS filesystem can only write about 550K of data per second. I was hoping to see 10 times that level of performance. :( I recall an async option to mount_nfs back in FreeBSD 2.1, but I can't find it documented anyhwere. Is there anything else I can do to improve NFS performance? -- Brian Tao (BT300, taob@risc.org) "Though this be madness, yet there is method in't"