From owner-freebsd-hackers Fri Jun 2 22:11:13 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id WAA28167 for hackers-outgoing; Fri, 2 Jun 1995 22:11:13 -0700 Received: from gndrsh.aac.dev.com (gndrsh.aac.dev.com [198.145.92.241]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id WAA28151 for ; Fri, 2 Jun 1995 22:11:05 -0700 Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id WAA09409; Fri, 2 Jun 1995 22:10:47 -0700 From: "Rodney W. Grimes" Message-Id: <199506030510.WAA09409@gndrsh.aac.dev.com> Subject: Re: A performance mystery To: henrich@crh.cl.msu.edu (Charles Henrich) Date: Fri, 2 Jun 1995 22:10:47 -0700 (PDT) Cc: freebsd-hackers@freebsd.org In-Reply-To: <199506030459.VAA27174@freefall.cdrom.com> from "Charles Henrich" at Jun 3, 95 00:59:13 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2145 Sender: hackers-owner@freebsd.org Precedence: bulk > > > We are not going to get the results I was looking for unless you > > are going to follow directions a *little* closer. > > > > ``repeat 10 iozone 16 8192'' > > Rod, you dont have to be so condescending (note, no expletives), I > dont *have* your original message. :-(. > > > You said this was a Pentium right?? If it is that number is low for > > a triton board: (ASUS PCI/I-P54TP4, 90Mhz CPU, 64MB memory) > > Its not a Triton. It was neptune right? Okay, then 18199013 / 1.30 == 13999240, about what I have seen for Neptune vs Triton chip set. > > IOZONE performance measurements: > > 1915685 bytes/second for writing the file > > 18199013 bytes/second for reading the file > > > > > 16 8192 3360694 2890287 ^^^^^^^ How come *this* number and the number below are different by a factor of 7??? It should have been the same. ``iozone 16 8192'' and the above line from ``iozone auto'' do exactly the same things yet we are seeing this *huge* descrepancy. > > Ding... very low memory bandwidth :-(. Makes for a very slow > > compile :-( > > Here is the results of the iozone, as ram-speed showed as well, the memory > bandwidth is very good: > > 2684354 bytes/second for writing the file > 14412641 bytes/second for reading the file But why is this so much different than what you gave me in the last message with iozone auto. The results should have been almost IDENTICAL. Infact that whole run of iozone auto showed the read numbers to *all* be very low for a Pentium class machine on the Compaq. Something is wrong here :-(. This is the results I get from the end of iozone auto, and from iozone 16 8192 on the same 64MB ASUS triton board: 16 8192 1927723 18199013 18199013 bytes/second for reading the file Notice the fact these numbers are *IDENTICAL*, but yours are not, again something does not add up :-( -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Custom computers for FreeBSD