Subject: Re: Enough for all my favorite OS's
From: civileme (civileme@mandrakesoft.com)
Date: Fri Jul 26 2002 - 06:48:05 AKDT
On Friday 26 July 2002 02:07 am, bryan@ak.net wrote:
> On Thu, Jul 25, 2002 at 10:10:52PM -0800, civileme
<civileme@mandrakesoft.com> wrote:
> > On Thursday 25 July 2002 11:43 pm, W.D.McKinney wrote:
> > > Now if my laptop could use one of these I'd be set :-)
> > > http://www.wdc.com/company/releases/Pressrelease.asp?release=1404
> > >
> > > /Dee
> >
> > Yep the only thing really wrong with it is that it is WD and it has the
> > same lack of ability on 57-byte CRCs as its predecessors.
>
> Er, sorry -- could you explain what that means?
Come to think of it, about a year ago I wrote a little routine called drakopt
because early kernel 2.4 stuff just wasn't doing very well on
idex=autotune
I had about 30 volunteer testers checking drakopt for bugs and performance,
and I remember one log quite well..
He had a WD drive on hdb and a Quantum on hda, and when a certain point was
reached in the testing of drive hda we had an error message
hdb: lost interrupt
Whoa! just cause I had upped the channel with
hdparm -c1 -d1 -m8 -a4 -u1 -X67
suddenly a request to hda was answered by hdb....
This sort of crosstalk on the ide channel is supposed to be impossible, yet
the timing of hdb permitted it.
The drive, a WD200BB, was moved to the hdc positionn and very interesting
things showed up from that....
data rate at -X68 (udma 4) with all other opts set well ..... 4.2Mb/s
data rate at -X66 (udma 2) witrh all other options optimal 9.6Mb/s
The tester then loaded RH 7.1 and got the astounding data rate of 1.62 Mb/s
Settings were not optimal for udma4 but it was at udma4 that the drive was
set.
Under interaction with me via email the tester tried udma2 with the setting
drakopt had found under mandrake 8.0 and again the performance was about 9.6
Mb/s
That is one of the reasons I wrote drakopt; because many drives often do
perform better at lower than their rated settings...
Maxtor 20s on a 1U server 28.95Mb/s at udma5 29.42Mb/s at udma4
Which would you choose? I set it at 66MHz rather than 100MHz because the
performance was (essentially) the same and there is better noise immunity at
the lower frequency (the total amount of charge flowing is a larger energy
signature in comparison to external noise).
Ah well, I am running off at the mouth. I do have a couple of WD drives
which I have been keeping around for software testing (they reliably throw
errors above 33Mhz, and that is what we want--our software should be able to
run on junk). But I would never buy one for any computer project.
Civileme
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
This archive was generated by hypermail 2a23 : Fri Jul 26 2002 - 06:48:08 AKDT