Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Expose proportional resident memory size #140

Closed
PierreF opened this issue Mar 26, 2020 · 3 comments · Fixed by #142
Closed

Expose proportional resident memory size #140

PierreF opened this issue Mar 26, 2020 · 3 comments · Fixed by #142

Comments

@PierreF
Copy link
Contributor

PierreF commented Mar 26, 2020

Currently the resident size is the sum of resident size from all process in the group, which will means a rather high value for processed that share significant amount of memory.

Linux do expose a proportional resident memory size in the smaps file, which is

The "proportional set size" (PSS) of a process is the count of pages it has
in memory, where each page is divided by the number of processes sharing it.

I think the sum of PSS is more interesting than sum of RSS, as RSS sum is way too high.

I'm willing to come with a PR, but that need update on procfs dependency and before working on this, I want to known your thought on whether this change should be done on your procfs fork or being submitted to prometheus/procfs and stop using the forked-procfs ? It seems that prometheus/procfs now has ProcStatus support which was the reason of the fork.

@flixr
Copy link
Contributor

flixr commented Mar 26, 2020

I'm also seeing this issue, but didn't quite spend enough time to find out where this is coming from...
But "correct" reporting of this would be very very welcome!

@PierreF
Copy link
Contributor Author

PierreF commented Apr 6, 2020

I'll start working on PRs to provide this by using prometheus/procfs, that is:

  • submit PR here to use prometheus/procfs instead of ncabatoff/procfs
  • submit PR on prometheus/procfs to export smaps information
  • submit PR here to using smaps information from prometheus/procfs. This PR will depend on the two above.

@PierreF
Copy link
Contributor Author

PierreF commented Apr 8, 2020

PRs created:

lporcheron pushed a commit to bleemeo/glouton that referenced this issue Jul 6, 2020
This use WIP as it depend on forks of process-exporter & procfs.
See ncabatoff/process-exporter#140
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants