Re: Interesting scheduling times - NOT

Peter T. Breuer (ptb@it.uc3m.es)
Fri, 25 Sep 1998 16:14:00 +0200 (MET DST)

"A month of sundays ago Neil Conway wrote:"
>
> Larry McVoy wrote:
> >
> > 2 7.58 (7.74 7.65 7.63 7.60 7.60 7.58 7.58 7.58 7.55 7.54 7.53)
> >
> > 2 7.73 (8.38 8.14 8.13 8.00 7.94 7.73 7.62 7.60 7.46 7.33 7.04)
> >
> Anybody else notice that Larry's results are monotonic on both sets of
> data ? The chances against monotonicity for each run are 2^9 to 1
> against.
>
> This says VERY loudly that these are not random differences - something

Quite correct. This is about as strongly correlated against run number
as you can get. Correcting for the deduced dependence on the run number
leaves a very small variance indeed in the rest. About 0.1 sd. But I
suspect Larry corrected for that also?

> And yeah, the median certainly covered *that* thing up, no?
>
> Comments?

It has always been necessary to gather the full run statistics. I don't
know why both parties keep going on about mins and medians ! (they're
interesting executive summary numbers sure, but we learn a whole lot
more from the distribution itself).

Peter

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