Date: Sat, 17 Dec 2011 11:23:02 -0500 From: Jason Hellenthal <jhell@DataIX.net> To: Doug Barton <dougb@freebsd.org> Cc: freebsd-stable@freebsd.org Subject: Re: swi4: clock taking 40% cpu?!? Message-ID: <20111217162302.GB3875@DataIX.net> In-Reply-To: <20111217162029.GA3875@DataIX.net> References: <4EEA5DD0.1040009@FreeBSD.org> <20111217162029.GA3875@DataIX.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Should also mention the kern.sched may be playing a part in this too.
On Sat, Dec 17, 2011 at 11:20:29AM -0500, Jason Hellenthal wrote:
>
>
> On Thu, Dec 15, 2011 at 12:51:28PM -0800, Doug Barton wrote:
> > Howdy,
> >
> > Web server under heavy'ish load (7 on a 2 cpu system) running
> > 8.2-RELEASE-p4 i386 I'm seeing this:
> >
> > PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU COMMAND
> > 12 root -32 - 0K 112K WAIT 0 129:01 39.99% {swi4: clock}
> >
> > Any ideas why the clock should be taking so much cpu? HZ=100 if that
> > makes a difference ...
> >
> >
>
> Without NTPd running test the following.
>
>
> apply "/usr/bin/time -ph sleep %1" 300 600 900
>
> If the results are skewed quite a bit then your system may benefit from a different HZ than what you have set. I have seen systems that require a HZ of 350 and as weird as it sounds NTPd may be tasting the clock too much just to try and keep time.
>
> --
> ;s =;
--
;s =;
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20111217162302.GB3875>