Ask stupid questions and you'll get a stupid answers, was: Technological advantages over Linux

[Steve O'Hara-Smith]

On Sun, 26 Jul 2020 09:57:44 -0400
Aryeh Friedman <aryeh.friedman at gmail.com> wrote:

> On Sun, Jul 26, 2020 at 9:55 AM Steve O'Hara-Smith <steve at sohara.org>
> wrote:
> 
> > On Sun, 26 Jul 2020 09:04:02 -0400
> > Aryeh Friedman <aryeh.friedman at gmail.com> wrote:
> >
> > > On Sun, Jul 26, 2020 at 8:43 AM Steve O'Hara-Smith <steve at sohara.org>
> > > wrote:
> > >
> > > >         Here is a variant of your experiment that should demonstrate
> > it.
> > > >
> > > > 1: Reboot machine, measure performance
> > > > 2: Memory stress machine to until swapping reduces performance
> > > > 3: Kill memory stressing process
> > > > 4: Disable swap - which forces all pages back into RAM
> > > > 5: Enable swap
> > > > 6: Loop to 2
> > > >
> > >
> > > Since stealing memory  from a running process that counts on it to be
> > > functional will crash the process and odds are that process is
> > > something low level and critical to keeping X running the above
> > > variant is not practical to do and thus my current solution has the
> > > same effect -- reboot.
> >
> >         It is very practical and it is the way I normally recover a
> > system from a memory usage spike, it works far better than just leaving
> > it alone. Nothing in that steals memory from running processes.
> >
> 
> Unless there is just not enough RAM to allow all the pages to be recovered
> which as far I can tell is the problem.

	If there is not enough RAM for the workload you are imposing then
that is your problem, but 24GB should be plenty for the workload you
describe absent a runaway. No system should be run at a level that requires
swap.

-- 
Steve O'Hara-Smith                          |   Directable Mirror Arrays
C:\>WIN                                     | A better way to focus the sun
The computer obeys and wins.                |    licences available see
You lose and Bill collects.                 |    http://www.sohara.org/