:: Re: [DNG] difficulty issuing comman…
Top Page
Delete this message
Reply to this message
Author: Arnt Karlsen
To: dng
Subject: Re: [DNG] difficulty issuing commands
On Sun, 1 Aug 2021 07:16:00 -0400, Haines wrote in message

> On Sun, Aug 01, 2021 at 02:54:55AM -0400, Steve Litt wrote:
> > Haines Brown said on Sat, 31 Jul 2021 21:54:56 -0400
> >
> > >Sorry for the ambivalent subject line.
> > >
> > >I'm running Beowulf with Fluxbox window manager but without a
> > >desktop environment. Without knowing what might have triggered it,
> > >I find it laborious to start some applications from CLI. What
> > >happens is that I get a hang (little spiral rotates) until I kill
> > >the process with Crtl-g and try again.
> >
> > Do you mean Ctrl-c ? I've never heard of ctrl-g killing commands.
> >
> > [snip other weird intermittent symptoms]
> >
> > I'd start by collecting data:
> >
> > df -h >> data.txt
> No problems here. No partition is more than 50%
> > vmstat -S M >> data.txt
> I don't know how to interpret all the info. But here it is:
> procs -----------memory---------- ---swap-- -----io----
> -system-------cpu----- r  b   swpd   free   buff  cache   si   so
> bi    bo   in   cs us sy id wa st 2  3      0    162   5852   6370
> 0    0    85    95    3    9 10  3 86  1  0

> Nothing here strike my eye
> > Also, run the htop command to see whether you're maxing out your
> > CPUs.
> Wow! Don't like this result, It looks like for the four cores, number
> is running 100%. The others are low. Currently my machine is running
> an automatic bacvkup. I do not have a swap partition. Mem is
> 3.7G/15.6G

..in my final Debian days I used 8GB swap on my 4GB ram laptops.
Upgrading to 8GB ram also had me raise that to 24GB swap on an
hunch there might be something clever about 32GB total memory,
rather than carrying on with the twice-ram-swap rule to 24GB.
Works wonderfully well for me.

..and I suspect some programs act up if they can't see any swap
memory at all, regardless of whether or not they actually need
any swap memory, all it really takes, are bad and badly hardcoded

> In the display of processes, CPU runs 98-101% I'm runnng two sessions
> of emacs on two virtyual desktops and it is the second that is
> hogging CPU.
> I see that in fiddling to start balance program, I left it in hung
> state (spinning spiral). So now use C-g to stop the process. This
> is enacs for stop the process. Ctl-c has no effect on the emacs
> process. I now look to see what effect stopping the process has.
> Sure enough. My CPU is back to normal. My backup running in
> the background calls occasionally for as much as 30%.
> So I'm back to the problem that I cannot run ~/.backup.bal without.

..what happens if you try run this with nice,
e.g. "nice -n 18 ~/.backup.bal &" ?

..or tie it to some specific cpu core(s)?
Etc for your Emacs sessions.

> It may be that this is a problem specific to emacs.
> I'll take the further steps if you think wise. It maey be that the
> problem with .backup.bal is an oddity. I'm about to reconstruct the
> bookmark. The problem with accessing bank not understanding my
> browser until I try to acess a second time is probably irrelevant.

..med vennlig hilsen = with Kind Regards from Arnt Karlsen
...with a number of polar bear hunters in his ancestry...
Scenarios always come in sets of three:
best case, worst case, and just in case.