Jack Stein wrote in a message to Roy J. Tellason:
JS> What are your cache settings for HPFS/FAT?
I'll have to go look...
(Took *forever* for things to come up...!)
All I can find in there with "cache" in it is this:
IFS=C:\OS2\HPFS.IFS /CACHE:512 /CRECL:4 /AUTOCHECK:CD
and this:
DISKCACHE=D,LW
which, if I'm remembering right, is saying that I'm only using 512k for
cache. That's not all that much, is it?
JS> I find most people in small (and large for that matter) memory
JS> machines have WAY too big a cache for OS/2. The quickest and
JS> easiest way to screw up an 8 meg system is wasting ram on a
JS> large cache.
I haven't messed with any of this stuff, at least not yet. All of what's in
the config.sys is whatever the install _put_ in there. Speaking of which,
there sure seems to be a lot of "stuff", lots of drivers for hardware I
don't have, etc. I wonder if these are at least part of why it's taking so
long for this machine to come up?
JS> I recommend you run CFGINFO4.ZIP and apply most of the settings
JS> he recommends for your system.
Hmm, funny you should mention that, as I just got it:
cfginfo4.zip 627944 19-Apr-98 23:31
(That time/date stamp is when it got here...)
In fact, I was going to post about such utilities, since there seem to be a
few of them out there, as well as text files that talk about sorting your
config.sys and such, and then you go ahead and tell me to get this one! :-)
Makes me feel like maybe I'm on the right track, or sumthin'...
(Added later) I did try it, followed their suggestions, and it _is_
perceptibly faster, though not by a whole lot.
JS> BTW, in my system, I noticed a bigger performace improvement
JS> when I added a new faster HD than I did when I added 12 megs of
JS> ram. Renumbering my message bases for example, went immediatly
JS> from 9-10 minutes to 3-4 minutes for 1500 *.msg's.
Hmm. Well, the drive I'm using in that machine is the 540M WD that used to
be in _this_ machine, and held my message bases (one hell of a lot more than
1500 of them) before I stuck that SCSI drive in here. It's not as fast / new
as either of the two drives I have in here now, but it's not too bad, I
guess.
I'll try that config program and see what happens.
JS> Remember, I ran my exact same 8 meg system under DV for 2 years
JS> before upgrading to OS/2, and most every area, performance
JS> either stayed the same or improved.
I would be interested in hearing more about how you made that particular
transition, since that's exactly what I'm looking at here. So far I've got
OS/2 versions of *some* of the software I run, but not all of it, not by a
long shot...
At least the bbs and tosser and front end (binkley) all come in OS/2 versions
and I have them all here!
WH> Seriously, an 8 meg machine with anything more than maybe
WH> (preferably not) the WPS running is about as much fun as
WH> practice bleeding. Same with the main machine: I was amazed
WH> at what 32 meg vs 16 did.
JS> I've heard this said many times by many people, but my
JS> experience is not CLOSE to this. If you are set up correctly,
JS> and running mostly DOS and OS/2 TEXT applications, 8 megs is
JS> very good, if not great.
Hmm. I haven't run any "TEXT" apps on there yet, just a few odds and ends
of what's already there either built in on the OS cdrom or from the Bonus
Pack, and mostly that's PM stuff. Except I stuck a copy of the OS/2 version
of InspectA on there.
JS> Disk thrashing shoul NOT occur, and if it does, something is
JS> wrong.
It sure seems to spend a lot of time hitting that drive anytime I do just
about *anything*. Takes a heck of a long time to boot, too, which is why I
snagged those config utilities in the first place.
JS> If you are hearing your HD "ticking" every few seconds, and
JS> things are dead slow, something is wrong.
Not "ticking", more of a "chatter" that goes on for several seconds. Boot
seems to take several _minutes_.
(Added later) The best way I can describe the noise is that it's the same
noise (same drive, too) that I used to get when I used that m'board to
access the 'net with typical windoze stuff about two or three years ago --
the fault was in the Netscape "cache" settings, I think, but no settings
seemed to do a whole lot of good, didn't improve it any, and the only way I
could stop the chatter was to disable the disk cache altogether.
JS> I've had that happen on rare occasions and never did figure out
JS> what caused it, but it was very unusual. I don't recall that
JS> happening since going to WARP a few years ago, but it may have
JS> stopped when I went to 20 megs... it's been awhile.
RJT> I forget now what it was that I did change, but it was some
RJT> of the other stuff you suggested...
JS> I made lots of tweaks to my system over the years, and when I
JS> got around to looking at CFGINFo4.ZIP, He had most everything
JS> that I had done as recommendations. I think we both got our
JS> info from the same sources, right here on FIDO.
The other thing I've been snagging is text files, both official-looking
stuff and FAQ/FGA files. There's still more to go yet, though.
JS> I might add that none of the "tweaks" I put in from the
JS> default set up changed any thing dramatically.
Well, I'll try sticking that config utility on there and see what happens.
email: roy.j.tellason%tanstaaf@frackit.com
---
---------------
* Origin: TANSTAAFL BBS 717-432-0764 (1:270/615)
|