Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!husc6!rutgers!lll-crg!hoptoad!rdm
From: rdm@hoptoad.uucp (Rich Morin)
Newsgroups: comp.arch
Subject: Re: I/O benchmarks
Message-ID: <1305@hoptoad.uucp>
Date: Mon, 17-Nov-86 17:32:28 EST
Article-I.D.: hoptoad.1305
Posted: Mon Nov 17 17:32:28 1986
Date-Received: Tue, 18-Nov-86 01:31:04 EST
References: <2657@princeton.UUCP>
Organization: Canta Forda Computer Laboratory
Lines: 22
Summary: watch out for buffering interactions

In article <2657@princeton.UUCP>, park@princeton.UUCP (Arvin Park) writes:
> 
>      I am assembling a collection of benchmarks to measure I/O system 
> performance.  ...  I welcome any contributions, suggestions, and words
> of wisdom or code if you have it.

Steve Wallach (of Convex) was interviewed in the November, 1985 issue of
UNIX Review.  He told the following, possibly relevant, story:

	... With all this physical memory {>100 MB}, we can make the
	disk cache as big as we like, so whenever we run up against
	I/O benchmarks, we just define a disk cache large enough to
	keep us from having to go out to disk.  As a result, our
	machines have screamed through benchmarks.  Some people cry,
	"Foul!  That's not a fair benchmark because I can't do that
	on my VAX" -- to which, of course, we respond, "Right".  Then
	we smile and don't say anything more.
-- 
Richard Morin, proprietor	{hoptoad,leadsv,lll-lcc}!cfcl!rdm
Canta Forda Computer Lab.	+1 415 994 6860
Post Office Box 1488		Full spectrum consulting services
Pacifica, CA  94044  USA	for science and engineering.