Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 3 Jul 2000 01:15:45 +0200 (CEST)
From:      Alban Hertroys <dalroi@wit401310.student.utwente.nl>
To:        Cy Schubert - ITSD Open Systems Group <Cy.Schubert@uumail.gov.bc.ca>
Cc:        freebsd-stable@FreeBSD.ORG
Subject:   Re: man nice(1) 
Message-ID:  <20000702231548.B23DD1F60@wit401310.student.utwente.nl>
In-Reply-To: <200007021418.e62EIww00785@cwsys.cwsent.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On  2 Jul, Cy Schubert - ITSD Open Systems Group wrote:
> In message <20000702120917.D5BA71E7E@wit401310.student.utwente.nl>, 
> Alban Hertroys writes:

> I see you're a csh user.  C Shell has a builtin nice which is 
> incompatible with /usr/bin/nice which is used by the Bourne shell and 
> any of its descendants.

tcsh, to be exact. I mentioned this problem to a couple of friends this
evening, and was pointed to this fact already. Sorry to waste your time
with that.


>> This is with a somewhat older 4.0-STABLE.
>> I found this when trying a buildworld "nicely", as my system hottens up
>> a tiny bit to much somewhere (probably my pci viper550 with bad fan)
>> causing spontaneous reboots after a couple of hours.
> 
> I don't see how the nice command would solve this.  It just assigns a 
> different priority to a command and its children causing them to be 
> selected for execution more or less frequently in relationship to other 
> processes running on the system.  If you have no higher priority work 
> on the machine, a lower nice value will not make any difference in 
> execution time or the amount of heat produced by your CPU.  Replacing 
> the bad fan or reducing your CPU's clock rate will help.

It did help, though, as my average load went from about 2 to 1. A
sustained load of 2 for a couple of hours seems to be a problem for my
system (which is due for replacement anyway), but a load of one is not
a problem as long as my room temperature is around 20 degrees.
It may also have helped that I didn't run X during today's buildworld
(which succeeded) if the problem is related to my viper550.

I realize this is not an optimal situation, but there is no real
solution to it until I pinpoint the heat source.

I also realize that nice values will only marginally help achieving
what I want. I have been toying with the idea to send a kill -STOP to
the group leader process every hour or so, and sent a kill -CONT a
quarter of an hour later to let things cool down again.
This is a bit trickier to achieve, though, as it involves timing when
the heat rices too much (depending on the environment temperature) and
'requires' checking the availability of the targetted process to not be
spammed by your own cron daemon (This is an old ASUS Pentium T2P4-board
without lm chips).

-- 

Alban Hertroys                              http://wit401310.student.utwente.nl
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
              The difference between your legs and your bicycle is
              that when they are in bad shape it is unwise to give
              your bicycle more training.



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000702231548.B23DD1F60>