[Bug 173541] High (0.60+, 1.00) idle load averages

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=173541

--- Comment #44 from Mark Millard <marklmi26-fbsd@yahoo.com> ---
(In reply to Stavros Filargyropoulos from comment #40)

Previously you listed (FreeBSD context):

128 threads:   5 running, 103 sleeping, 20 waiting

vs. for the OpenBSD context now:

51 threads: 50 idle, 1 on processor

Focusing just on the thread counts (the other figures
are not likely to be as stable of figures) . . .

I can not tell how much of 128 vs. 50 is a mismatched
context/configuration vs. how much is differences in
software design/implementation. Whatever the mix, the
128 vs. 50 is notable all by itself.

That actually points out an issue for the 4 core
test: some types of thread counts do scale with core
counts in order to be able to put more cores to use
if needed. (It does not mean that they are busy.)
I'm not sure how similar the thread counts would
be. I might need to stop nfs or some such to have
a more similar context relative to thread counts.

-- 
You are receiving this mail because:
You are the assignee for the bug.