Subject: Re: Load average calucation under NetBSD 2.0_BETA
To: None <netbsd-users@NetBSD.org>
From: Ignatios Souvatzis <ignatios@cs.uni-bonn.de>
List: netbsd-users
Date: 06/09/2004 11:49:54
--XvKFcGCOAo53UbWW
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Wed, Jun 09, 2004 at 09:39:57AM +0000, Matthias Scheler wrote:
> In article <20040609093416.GD23269@drowsy.duskware.de>,
> 	Martin Husemann <martin@duskware.de> writes:
> > On Wed, Jun 09, 2004 at 09:19:22AM +0000, Matthias Scheler wrote:
> >> load averages:  1.43,  1.27,  1.17                                    =
 11:16:43
> >> 123 processes: 2 runnable, 120 sleeping, 1 on processor
> > 2 runnable processes - so expect some load.
>=20
> But the CPU was 99% idle.
>=20
> > For whatever value of "properly".
>=20
> I wouldn't expect a system idleing more than 90% of the time to report
> a load above 1.

Does your system have a seperate, mistuned statistics clock interupt? Else
a process that wakes up periodically to go to sleep again after a short time
can add substantial displayed load (1.0 more than the real one) - without
actually loading the machine hard.

Some versions of Ultrix used to show that behaviour excessively - I've=20
seen xload baselines of up to 4.0 .

Regards
	-is

--XvKFcGCOAo53UbWW
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: 2.6.i

iQEVAgUBQMbdPzCn4om+4LhpAQEXgwf5AfRDF1Gq3EvIjuw0j0Ew3G1RaHseJrQN
JSs+F3rqSkIUm65gq2Fe9NRnnW/Mb2pil3975uMi9ZYeij6rCdljAuZXdffgwglK
DEZy9GvLmAm641fD4h3O2r3oiBLlxgekztbBB50iBlCwrg1djq6ujtwQtoky6mis
HvCKKnRrodN/zFR2joKabwClzoulR1gljFNtn/rqaV9jbUXJAk0A0hbyRKdc1K5J
HLfRPgDc3ksLXcN3kwoOcE6ekdbzmzIdfxolKmHtdTA6+473U6baabxqkIg5LZx1
1DZrGNR9jHgQW3jH+64fmOZwdRbzhQlqaNL/OQB0Eu8xGy4baEFPPA==
=g+xM
-----END PGP SIGNATURE-----

--XvKFcGCOAo53UbWW--