Thread: Jan's benchmark suite and Matlab

Jan's benchmark suite and Matlab

From
Steve Bergman
Date:
I've successfully completed a preliminary "orientation" run of Jan's
tpcw suite.  And now I'm wondering what to do with the resulting .out
file.  The docs say that it is a matlab function.  So do you have to
have matlab to make sense of the results?  Should Octave work? (I'm not
very familiar with either tool.)  Or is there meaningful info to be
gleaned with only the naked eye?

FWIW, here is the tail end of the file:

dat.numinteractions = 21338;
dat.startRU =  1091907248390;
dat.startMI =  1091907548390;
dat.startRD =  1091908748390;
dat.term    =  1091908808390;
dat.slowDown = 1.0;
% Errors
% Total Errors: 0


I understand that RU is the ramp up period, RD is ramp down, and MI is
measurement interval, and and numbers seem to be time stamps.

So is dat.numinteractions something meaningful?

Thanks,
Steve


Re: Jan's benchmark suite and Matlab

From
Alvaro Herrera
Date:
On Sat, Aug 07, 2004 at 06:33:52PM -0500, Steve Bergman wrote:

> FWIW, here is the tail end of the file:
>
> dat.numinteractions = 21338;
> dat.startRU =  1091907248390;
> dat.startMI =  1091907548390;
> dat.startRD =  1091908748390;
> dat.term    =  1091908808390;
> dat.slowDown = 1.0;
> % Errors
> % Total Errors: 0
>
> I understand that RU is the ramp up period, RD is ramp down, and MI is
> measurement interval, and and numbers seem to be time stamps.

Apparently this is in milliseconds, which would mean that your rampup
time was 5 minutes.  Jan once said to me on IRC that you should leave at
least half an hour on rampup, so that the cache structures (ARC buffers)
have time to get useful numbers.

--
Alvaro Herrera (<alvherre[a]dcc.uchile.cl>)
"Estoy de acuerdo contigo en que la verdad absoluta no existe...
El problema es que la mentira sí existe y tu estás mintiendo" (G. Lama)