*BSD News Article 21350


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!network.ucsd.edu!ogicse!netnews.nwnet.net!ns1.nodak.edu!agassiz.cas.und.NoDak.Edu!agassiz.cas.und.NoDak.Edu!rand
From: rand@cs.UND.NoDak.Edu (Douglas K. Rand)
Newsgroups: comp.os.386bsd.questions
Subject: Re: VLB IDE cache controller
Message-ID: <RAND.93Sep22161811@agassiz.cas.und.NoDak.Edu>
Date: 22 Sep 93 11:18:13 GMT
Article-I.D.: agassiz.RAND.93Sep22161811
References: <27mf1g$ajt@news.u.washington.edu> <RAM.93Sep21145412@xor.epi.wisc.edu>
Organization: University of North Dakota, Grand Forks, ND
Lines: 36
NNTP-Posting-Host: agassiz.cas.und.nodak.edu
In-reply-to: ram@xor.epi.wisc.edu's message of 21 Sep 1993 19:54:12 GMT

** <yaot@hardy.u.washington.edu> writes:

>>> I just upgraded my old ISA IDE controller to a VLB cache version.
>>> However, after some thorough test, I notice that the cache
>>> controller doesn't seem to improve the I/O performance at all.

** In article <RAM.93Sep21145412@xor.epi.wisc.edu>,
   ram@xor.epi.wisc.edu (Ram Bhamidipaty) writes: 

Ram> I would expect almost no benefit from a caching controller, since
Ram> the cache system in 386bsd is already doing that job.  [...] but
Ram> an IDE controller can only handle two disks, which can be handled
Ram> by the ISA bus.

I've got a Promise Technology VLB caching controller with 4 Mbytes on
the controller (20 Mbytes main memory) and I see a signfigant
difference with the cache. Here are some simple performance numbers
for just the linking phase of building a new kernel:

                        First time      Second time
cache disabled:		16.41 secs	16.21 secs
cache enabled:		 9.20 secs	 3.28 secs

What I did is time make, and then remove netbsd, and then time the
make again. With the cache disabled the second time is almost the same
as the first (.2 secs difference). With the cache enabled the time is
almost cut by 2/3. Also note that even the first time with the cache
is faster by a little more than 40%.

I'm not advocating putting a large disk cache in your UNIX boxes, but
it does help mine. (It *really* helps for DOS/Windows!)
--
Douglas K. Rand                     UND Aerospace - Scientific Computing Center
Home:   +1 218 773 0120                              University of North Dakota
Office: +1 701 777 2801                    Box 9022, Grand Forks ND  58202-9022
Internet: rand@cs.UND.NoDak.Edu             UUCP: ...!uunet!plains!agassiz!rand