Re: Fwd: Re: [patch][rfc] quell interactive feeding frenzy

[Al Boldi <a1426z@gawab.com>]

Con Kolivas wrote:
> Hi Al

Hi Con!

> On Tuesday 11 April 2006 00:43, Al Boldi wrote:
> > After that the loadavg starts to wrap.
> > And even then it is possible to login.
> > And that's not with the default 2.6 scheduler, but rather w/ spa.
>
> Since you seem to use plugsched, I wonder if you could tell me how does
> current staircase perform with a load like that?

With plugsched-2.6.16 your staircase sched reaches about 40 then slows down, 
maxing around 100.  Setting sched_compute=1 causes console lock-ups.

With staircase14.2-test3 it reaches around 300 then slows down, halting at 
around 500.

Your scheduler seems to be tuned for single-user multi-tasking, i.e. 
concurrent tasks around 10, where its aggressive nature is sustained by a 
short run-queue.  Once you go above 50, this aggressiveness starts to  
express itself as very jumpy.

This is of course very cpu/mem/ctxt dependent and it would be great, if your 
scheduler could maybe do some simple on-the-fly benchmarking as it 
reschedules, thus adjusting this aggressiveness depending on its 
sustainability.

Thanks!

--
Al

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