Re: osf/1 slowness

Justin Dolske (dolske@cis.ohio-state.edu)
Wed, 16 Apr 1997 15:24:42 -0400 (EDT)


On Wed, 16 Apr 1997, Me wrote:

> I know that a pentium 100 is not faster than a DEC,

Beware of broad generaliztions. DEC makes a wide variety of platforms. I
would not be surprised if low-end workstations perform much worse than a
Pentium. Also, beware of the *age* of the machines you're using. Because
technology moves so fast, those $100,000 boxes a university bought a few
years ago are probably considered slow now.

> One more thing,
> I think the DEC client is set to either nice -20 or nice -10, that could
> make a difference...

Absolutely. If people are running background jobs (they don't have to be
logged in), priority will likely be going to those jobs (DESCHALL will
still get some work done, though).

Play with "ps" to see what's running. An easier approach is to lookk at
the output of "uptime":

3:22pm up 22:50, 4 users, load average: 0.08, 0.09, 0.18

The first number indicates the number of processes currently competing for
CPU time (the other two nuber are averages over longer periods of time).

Near-zero values, like here, indicate the system is basically idle.

> Oh, does anyone know what the DEC clients command line parameters
> are, or how thier used, can you make it hog up more prossesor time?

-m does a performance measurement. The Other options are non-functional or
broken, don't worry about them.

Justin Dolske <URL:http://www.cis.ohio-state.edu/~dolske/>
(dolske@cis.ohio-state.edu)
Graduate Fellow / Research Associate at The Ohio State University, CIS Dept.
-=-=-=-=-=-=-=-=-=-=-=-=-=- Random Sig-o-Matic (tm) -=-=-=-=-=-=-=-=-=-=-=-=-
Some mornings I wonder if it's worth chewing through the
leather restraining straps.