*BSD News Article 51615


Return to BSD News archive

Newsgroups: comp.unix.bsd.freebsd.misc
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!news.kei.com!newshost.marcam.com!usc!howland.reston.ans.net!torn!watserv2.uwaterloo.ca!undergrad.math.uwaterloo.ca!bcrwhims
From: bcrwhims@undergrad.math.uwaterloo.ca (Carsten Whimster)
Subject: Re: SCSI - throughput
Message-ID: <DFF114.2n2@undergrad.math.uwaterloo.ca>
Sender: news@undergrad.math.uwaterloo.ca (news spool owner)
Date: Sun, 24 Sep 1995 15:15:04 GMT
References: <43bufs$ekb@nx2.hrz.uni-dortmund.de> <1995Sep21.090850.12239@wavehh.hanse.de> <DFBBvL.4tJ@undergrad.math.uwaterloo.ca> <MICHAELV.95Sep23000036@mindbender.headcandy.com>
Nntp-Posting-Host: cayley.uwaterloo.ca
Organization: University of Waterloo
Lines: 29

In article <MICHAELV.95Sep23000036@mindbender.headcandy.com>,
Michael L. VanLoon <michaelv@MindBender.HeadCandy.com> wrote:
|  In article <DFBBvL.4tJ@undergrad.math.uwaterloo.ca> bcrwhims@undergrad.math.uwaterloo.ca (Carsten Whimster) writes:
|  
|     Also, what are some decent cached SCSI cards? I have an old VLB
|     Promise IDE 4Mb cached controller, and I want to exceed the
|     performance of this piece before I buy.
|  
|  You don't want a caching card.  You'll just be spending your money on
|  nothing useful.  The caching in a BSD unix is much more sophisticated
|  and highly effective already.  If you couple that with a decent fully
|  bus-mastering SCSI card, you should have the ideal setup.
|  
|  Caching controllers are of questionable value even in DOS systems
|  (software caching is reported to be just as fast, and significantly
|  cheaper).  A modern unix works even better, because it can reallocate
|  memory pages intelligently and pretty much nullify the effects of the
|  hardware cache on the card.

For whatever reason, it does work well on OS/2, though. I had it in,
and got very nice performance. I recently had to take it out due to a
conflict, and I am now using the EIDE on my motherboard, and it is
_noticeably_ slower doing certain things. For this reason I presumed
that it is a good thing :)
-- 
          Carsten Whimster     bcrwhims@uwaterloo.ca
           EDM/2 Associate Editor and Book reviewer
            Team OS/2           POV-Panel/2 author
       http://www.undergrad.math.uwaterloo.ca/~bcrwhims/