[PLUG] Insights for the non-SysAdmins among us

Rich Shepard rshepard at appl-ecosys.com
Tue Oct 2 08:43:00 PDT 2018

   There are a few of us who are not computer professionals and some might
benefit from what I discovered about my desktop server/workstation.

   My replacement desktop (still being set up) is in a case lacking a hard
drive activity light, which turns out to be a Good Thing(TM). While I was
looking for a software solution Ed Koenig reminded me of gkrellm. I invoked
that on the replacement and learned how to configure it to suit my needs.
Last evening I watched a while while the system built > 100 R packages and
saw that the composite CPU usage (of 16 threads) remained a constant 6%.

   This prompted me to invoke it on the old desktop (dual core Athlon II
CPU), too, which I did this morning. I wondered why CPU 0 was steady on
99-100% and CPU 1 would occasionally show activity. Running top found the
PID of the daemon consuming all that processor core, and 'ps ax | grep PID'
revealed it to be atd, the tool that runs jobs at specified times.

   The atq command showed four jobs running, a couple more than once, and all
from Nov. 3rd and 10th of 2017! What those jobs were I have no idea, nor do
I know why they have been running for 11 months. So I removed them (atrm)
and CPU 0 usage hit the floor and remains docile.

   Now I understand why CPU temperatures have been higher than they were in
the past and why some models run very slowly or overheat the CPUs. My
intention is to make others like me aware of the benefits of constant
monitoring of CPU and disk activities using gkrellm. We don't know when
something is happening that needs attention.



More information about the PLUG mailing list