4

enter image description here

I have read that the Green bar represents the normal user process and the Red bar represents the time spent in kernel(kernel, iowait etc).

I'm running a jupyter-notebook that runs n=-1 jobs, and by definition all my cores should be used as normal user process and the htop should show high per core usage. And the normal behaviour while running this notebook was that the average per core usage used to hit 98-99%.

But weirdly now running the same notebook, the per core usage is limited to a max of 18% with an abrupt increase in kernel time. I want to understand why this is happening.

BlimBlam
  • 336
  • 3
  • 11
  • 1
    What did change from "earlier" to "now"? Please [edit] your question to addthis information, don't use comments to answer. – Bodo Sep 22 '20 at 16:01
  • Without a way to reproduce the issue, there's no way to answer the question. // but a guess is that either the jobs stops very quickly (so there are a lot of process start/stop overhead) or it's I/O bound or something. – user202729 Dec 28 '20 at 10:29

0 Answers0