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

activate pcp archive compression #15

Open
fche opened this issue Jan 23, 2018 · 6 comments
Open

activate pcp archive compression #15

fche opened this issue Jan 23, 2018 · 6 comments

Comments

@fche
Copy link
Collaborator

fche commented Jan 23, 2018

This should in theory reduce our PV disk space requirements by a factor greater than 10, though in exchange for more temporary RAM. performancecopilot/pcp#422 performancecopilot/pcp#386

@fche fche added this to the h milestone Jan 23, 2018
@fche
Copy link
Collaborator Author

fche commented Jan 24, 2018

This also requires pmwebd side changes to process compressed archives properly (looking for .xz file extensions, filename generation during ac_refresh timestamping).

@fche
Copy link
Collaborator Author

fche commented Mar 26, 2018

also previously needed:

  • testing to confirm that memory-rlimit forced fallback for decompression works in libpcp; i.e., try to keep going with /tmp-based decompression if -ENOMEM
  • even better, permit a decompression-buffer memory budget to be set ahead of time (env var?), so we throttle on-the-fly stuff before hitting kernel-level ENOMEM

@fche fche added the blocked label Mar 26, 2018
@goodwinos
Copy link
Collaborator

@fche is this still a blocker for OSIO? In any case - before embarking on this feature, you mentioned you were going to manually compress a largish archive collection and see how well pmwebd works with transparent decompression (for both memory and CPU resource usage)

@goodwinos
Copy link
Collaborator

@fche what was are your conclusions after testing? IIRC we discussed this a while back and tentatively concluded the overheads were too high for pmwebd, especially on startup. Is that correct?

@fche
Copy link
Collaborator Author

fche commented Oct 2, 2018

The overheads -are- high. But since the need still exists to ease disk usage and/or to allow longer term retention, the capability is still needed.

@fche
Copy link
Collaborator Author

fche commented Oct 2, 2018

The overheads -are- high. But since the need still exists to ease disk usage and/or to allow longer term retention, the capability is still needed.

By the way, one way to reduce that overhead somewhat is to keep archives open longer. https://bugzilla.redhat.com/show_bug.cgi?id=1619708#c9

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants