[arm-allstar] High CPU

[Bryan St Clair]

Thanks for the replies, all.

The reason I emailed the question was I had my home node become
unresponsive via ssh and http last night and it had never happened before.
Noticed the load times and the question was born.

Did a rf reboot and seemed like all was well, I just started to watch those
load times.

Thanks for everyone's reply.

Bryan

On Fri, Jan 24, 2020, 08:21 "Doug Crompton via ARM-allstar" <
arm-allstar at hamvoip.org> wrote:

> Just to add to that as Dave and I have talked about this for several years
> now.  I would basically disregard load averages on the Pi. They are not
> representative of actual CPU loading. The best way to determine CPU is
> loading is to watch the temperature using  cpu_stats.sh  Assuming you have
> stable ambient temperatures where the Pi is located the changes in CPU
> temperature are an indication of load. The Hamvoip code is really not very
> stressful on the CPU even with large numbers of connections. If you really
> want to see load some of the test programs will put the CPU into a
> frequency fold back state with temperatures over 85C! Assuming a 72F
> ambient a Pi with heatsinks and a fan should never go out of the the green
> region running Allstar. My heavily loaded Pi 4 turbo mode hubs run in the
> low 40's C all the time. Of course if your ambient is higher then you would
> likely see higher CPU temperatures.
>
>
> *73 Doug*
>
> *WA3DSP*
>
> *http://www.crompton.com/hamradio <http://www.crompton.com/hamradio>*
>
> On Fri, Jan 24, 2020 at 5:17 AM "David McGough via ARM-allstar" <
> arm-allstar at hamvoip.org> wrote:
>
> >
> > Hi Bryan,
> >
> > The answer to this question is simply that the "load average" numbers
> show
> > more that just CPU use. Under Linux, the load average is a combined
> number
> > including CPU usage, I/O usage, VM wait conditions (like swapping) and
> > time spent where a program is waiting on system resources (like
> > uninterpretable wait conditions), etc. It's easy to produce a program
> that
> > generates a 1.0 load average, while using virtually no CPU time or other
> > system resources.
> >
> > The vmstat program (man vmstat) is one which will probably give you a
> > better picture of actual system utilization.
> >
> > And, I just googl'ed the subject and found some possibly useful web pages
> > with more in-depth discussion:
> >
> >
> >
> https://unix.stackexchange.com/questions/393566/what-does-proc-loadavg-s-cpu-and-io-utilization-mean
> >
> > http://www.brendangregg.com/blog/2017-08-08/linux-load-averages.html
> >
> >
> > 73, David KB4FXC
> >
> >
> >
> >
> > On Thu, 23 Jan 2020, "Bryan St Clair via ARM-allstar" wrote:
> >
> > > All,
> > >
> > > I have noticed a higher than normal CPU load on my Pi.  Running 1 radio
> > and
> > > a 2nd radioless setup, streaming too.
> > >
> > > Loads are always in the .30+ range and I have seen them in the .70+ a
> few
> > > times.  Maybe I'm getting old, but I don't really remember them over
> even
> > > .10 most days.
> > >
> > > Pi 3B, no real changes of recent, updates performed this week, current
> > > build.
> > >
> > > What CPU loads are you seeing? I was checking on CLI.
> > >
> > >
> >
> > _______________________________________________
> >
> > ARM-allstar mailing list
> > ARM-allstar at hamvoip.org
> > http://lists.hamvoip.org/cgi-bin/mailman/listinfo/arm-allstar
> >
> > Visit the BBB and RPi2/3/4 web page - http://hamvoip.org
> >
> _______________________________________________
>
> ARM-allstar mailing list
> ARM-allstar at hamvoip.org
> http://lists.hamvoip.org/cgi-bin/mailman/listinfo/arm-allstar
>
> Visit the BBB and RPi2/3/4 web page - http://hamvoip.org
>