Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!zaphod.mps.ohio-state.edu!rpi!leah!emb978
From: emb978@leah.Albany.Edu (Eric M. Boehm)
Newsgroups: comp.sys.ibm.pc
Subject: Additional information on how to determine cache size
Summary: Additional information
Keywords: disk cache
Message-ID: <2527@leah.Albany.Edu>
Date: 11 Feb 90 16:35:52 GMT
References: <2525@leah.Albany.Edu>
Distribution: usa
Organization: The University at Albany, Computer Services Center
Lines: 23

I neglected to include some pertinent information in my previous post.
My hard disk is a 160 MB 16.5 MS ESDI Microscience 2160. It is split
into five logical drives (c-f == 32 MB, g == 20MB). The controller is a
Western Digital WD1007A-WA2 with a 32KB cache built into the controller
card. 

Buffers is currently 20, files=40. I do a lot of compilation (usually
move the code to the ramdisk and compile there). Edit small files (<
64KB), use some database stuff with a large file (around 400 KB). I also
use Windows 386 (but I changed the config.sys to use buffers=10,
smartdrv uses 384KB). In the future, I envision using something like
MUSH or PC-MAIL and possibly Desqview. I also do a lot of
telecommunications.

I have had suggestions to try different configurations, which I have
done to a small extent. I am looking for ways to more rigorously measure
performance changes than something like "X *seems* slower" or "Y *seems*
faster". In other words, an objective way to measure the changes.

-- 
Eric M. Boehm
EMB978@leah.Albany.EDU
EMB978@ALBNYVMS.BITNET