I'm having a server running around 250 docker containers. Once they are all up and running, htop shows a 32-33% RAM usage as shown below: enter image description here

I've also installed netdata for a more granular monitoring, but then I get the following (after a restart and ramp up of the containers): enter image description here

From what I see in netdata there are about 20Gb of RAM which are marked as cached, however htop does not show them?

I had several errors in which my node process monitoring the containers is running out of heap space (you can actually see that moment in the netdata graph when the measurements are not recorded for a small period).

What is this cached memory and why is it taking 20Gb? I would believe that my node process heap space is limited to 1Gb.

free -g confirms what netdata sees: enter image description here

Is there a way to know which process is using that memory?

  • Please post text, not screenshots. Not everybody can see images, text from images can't be copied, search engines can't process them, and there may be other reasons why this practice is frowned upon. Having said that, htop seems to report the used figure from free. It also seems to show cached memory in yellow. cached is RAM used for caching disk blocks, with the purpose of speeding up disk accesses. Cached memory is not associated with a single process, and sometimes with no process at all. Aug 7, 2021 at 2:06
  • linuxatemyram.com Aug 10, 2021 at 23:25


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.