From owner-freebsd-hackers Fri Apr 18 10:47:47 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA21509 for hackers-outgoing; Fri, 18 Apr 1997 10:47:47 -0700 (PDT) Received: from uno.sat.t.u-tokyo.ac.jp (uno.sat.t.u-tokyo.ac.jp [133.11.70.160]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA21490 for ; Fri, 18 Apr 1997 10:47:29 -0700 (PDT) Received: by uno.sat.t.u-tokyo.ac.jp (8.8.5/8.7.3) with ESMTP id CAA09011; Sat, 19 Apr 1997 02:47:23 +0900 (JST) To: Jaye Mathisen Cc: hackers@freebsd.org Subject: Re: Mildly interesting NFS numbers. In-Reply-To: Your message of "Tue, 1 Apr 1997 22:05:42 -0800 (PST)" References: X-Mailer: Mew version 1.69 on Emacs 19.28.1 / Mule 2.3 Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sat, 19 Apr 1997 02:47:23 +0900 Message-ID: <9009.861385643@sat.t.u-tokyo.ac.jp> From: Hidetoshi Shimokawa Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk From: Jaye Mathisen Subject: Mildly interesting NFS numbers. Date: Tue, 1 Apr 1997 22:05:42 -0800 (PST) Message-ID: mrcpu> mrcpu> Client is a Digital Prioris HX6000, P6-200/512cache, 32MB RAM, SMC mrcpu> 100Mbit ethernet. mrcpu> mrcpu> Server is a Network Appliance F540, 100GB disk, 8MB NVRAM, 128MB cache, mrcpu> 100MBit ethernet. mrcpu> mrcpu> I tried both TCP and UDP mounts, version 3 and version 2 NFS. The mrcpu> best performance was UDP version 3, for what that's worth. mrcpu> mrcpu> # iozone 500 mrcpu> Writing the 500 Megabyte file, 'iozone.tmp'...86.304688 seconds mrcpu> Reading the file...90.789062 seconds mrcpu> mrcpu> IOZONE performance measurements: mrcpu> 6074849 bytes/second for writing the file mrcpu> 5774792 bytes/second for reading the file mrcpu> mrcpu> using 64k block sizes resulted in about a 4% perf increase, however, it mrcpu> dropped CPU utilization on the Netapp from 100% to about 65%. Great. Netapp seems very good server, but it's too expensive :-( BTW, how many nfsiods did you run? If you ran only 4 nfsiod, I guess you could get better number with running 8 or more iods. /\ Hidetoshi Shimokawa \/ simokawa@sat.t.u-tokyo.ac.jp PGP public key: finger -l simokawa@sat.t.u-tokyo.ac.jp