lkml.org 
[lkml]   [2000]   [Oct]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] VM fix for 2.4.0-test9 & OOM handler
Rik van Riel wrote:

> On Mon, 9 Oct 2000, Marco Colombo wrote:
> > On Fri, 6 Oct 2000, Rik van Riel wrote:
> >
> > [...]
> > > They are niced because the user thinks them a bit less
> > > important.
> >
> > Please don't, this assumption is quite wrong. I use nice just to
> > be 'nice' to other users. I can run my *important* CPU hog
> > simulation nice +10 in order to let other people get more CPU
> > when the need it.
>
> In that case the time the process has been running and the
> CPU time used will save the process if it's been running for
> a long time.
>
> Please read the /entire/ algorithm before making rash
> conclusions like this.
>
> If nice is used for important, long-running tasks, the fact
> that they are long-running will save them (and be honest,
> would you really care if a simulation would be killed after
> 5 minutes? it's only inconvenient if it gets killed after
> a few hours...)

Main point is that in many enviroments the niced processes
are the ones the system exists for in the first place.
I.e they are the most important ones.
It would be completely unacceptable, say in research enviroment I know,
if my computational job will be killed because somebody launches Netscape or
some interactive game. The OS which does this is then will not be considered
suitable
for research.

Niced processes are usually the most complex one organizationally, and quite
often
are fairly complex to launch (not just click a button on your Gnome panel).
Moreover background niced processes by definition
are typically run unattended so need to relaunch has yet to be detected.
I, for example, typically lauch sequence of simulations
simultaniously on 20 Alphas running Linux, which then run for 3-4 days.
So yes, I do care that simulations are not killed after 5 minutes.
because, first I have to notice this (or write and run monitoring program),
second I have
to find what was exactly run on this specific machine which was killed and so
on.


>
>
> > But if you put the logic "niced == not important" somewhere into
> > the kernel, nobody will use nice anymore. I'd rather give a
> > bonus to niced processes.
>
> This doesn't make ANY sense at all. The objective is to destroy
> the least amount of work, which means giving a bonus to processes
> which have used a lot of CPU time already ... regardless of nice
> value.
>

Well, at the given CPU time spent, the niced process used much more real
time
than unniced. So the loss in terms of human hours is larger when it is
killed.
It shows quickly when you have some deadline to meet. Believe it is not
pleasant
to come in the morning and find the simulations you launched overnight
disappeared.
The day you planned on spending analysis the output results is gone.



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 12:39    [W:0.141 / U:2.344 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site