Skip to content

osd: optimize PG removal (part1)#37314

Merged
tchaikov merged 6 commits intoceph:masterfrom
ifed01:wip-ifed-faster-rm-p1
Dec 5, 2020
Merged

osd: optimize PG removal (part1)#37314
tchaikov merged 6 commits intoceph:masterfrom
ifed01:wip-ifed-faster-rm-p1

Conversation

@ifed01
Copy link
Contributor

@ifed01 ifed01 commented Sep 22, 2020

This is the first set of patches intended to improve user experience while doing pool/PG removal.
Includes previous position reuse for collection listing, trimming removed object from bluestore cache (which speeds-up final collection reaping) plus some means to gather statistics on the removal process.

Fixes: https://tracker.ceph.com/issues/47044

Signed-off-by: Igor Fedotov ifedotov@suse.com

Checklist

  • References tracker ticket
  • Updates documentation if necessary
  • Includes tests for new functionality or reproducer for bug

Show available Jenkins commands
  • jenkins retest this please
  • jenkins test classic perf
  • jenkins test crimson perf
  • jenkins test signed
  • jenkins test make check
  • jenkins test make check arm64
  • jenkins test submodules
  • jenkins test dashboard
  • jenkins test api
  • jenkins test docs
  • jenkins render docs
  • jenkins test ceph-volume all
  • jenkins test ceph-volume tox

@tchaikov

This comment has been minimized.

@ifed01 ifed01 force-pushed the wip-ifed-faster-rm-p1 branch from 1560978 to e813f64 Compare September 23, 2020 11:08
@ifed01 ifed01 requested a review from a team as a code owner September 23, 2020 11:08
@ifed01 ifed01 force-pushed the wip-ifed-faster-rm-p1 branch from e813f64 to 172cf71 Compare September 23, 2020 12:33
@ifed01
Copy link
Contributor Author

ifed01 commented Sep 23, 2020

this change breaks crimson build, see

@tchaikov - fixed, thanks!

Copy link
Member

@neha-ojha neha-ojha left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ifed01 could you please attach your performance test results to this PR?

@ifed01
Copy link
Contributor Author

ifed01 commented Sep 29, 2020

@ifed01 ifed01 force-pushed the wip-ifed-faster-rm-p1 branch from 172cf71 to 66317cd Compare September 29, 2020 18:04
@ifed01
Copy link
Contributor Author

ifed01 commented Sep 29, 2020

Here is another set of numbers for PG removal;
https://docs.google.com/spreadsheets/d/17V2mXUDEMAFVmSC67o1rQtrWNnAm_itBgzdxh1vRJMY/edit?usp=sharing

Columns B-E at "Overall" sheet cover both original performance and the changes from this PR.

@ifed01
Copy link
Contributor Author

ifed01 commented Sep 30, 2020

jenkins test classic perf

@ifed01
Copy link
Contributor Author

ifed01 commented Sep 30, 2020

jenkins test classic perf

@ifed01 ifed01 force-pushed the wip-ifed-faster-rm-p1 branch from b4b4d3c to 66317cd Compare September 30, 2020 18:15
@ifed01
Copy link
Contributor Author

ifed01 commented Oct 1, 2020

jenkins test make check please

@ifed01
Copy link
Contributor Author

ifed01 commented Oct 1, 2020

jenkins test make check

1 similar comment
@ifed01
Copy link
Contributor Author

ifed01 commented Oct 1, 2020

jenkins test make check

@palza
Copy link

palza commented Oct 18, 2020

hi team
is there any update on completion of review and merging pr?
the most important for me is 30bfcf8

Copy link
Member

@neha-ojha neha-ojha left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this looks good to me overall, @aclamk do you mind taking a look at the bluestore changes?
I think we should merge this PR first and keep the other improvements in #37496.

@ifed01
Copy link
Contributor Author

ifed01 commented Nov 16, 2020

@aclamk - I've resolved most of your comments but trim-on-remove logics simplification. The latter seems not working to me, see my relevant comment...
So please give this PR another try. Thanks!

@tchaikov
Copy link
Contributor

jenkins test make check

Signed-off-by: Igor Fedotov <ifedotov@suse.com>
Signed-off-by: Igor Fedotov <ifedotov@suse.com>
Signed-off-by: Igor Fedotov <ifedotov@suse.com>
Signed-off-by: Igor Fedotov <ifedotov@suse.com>
1) Do not differentiate all-flash and hybrid(ssd+hdd) deployments since
they both bound to SSD drive backing DB volume.
2) Drop osd_delete_sleep_hybrid from 2s to 1s
3) Raise osd_delete_sleep_ssd to 1s since 0 is a way too low - having
it that low tend to cause a significant negative imact on OSD's regular
operation.

Signed-off-by: Igor Fedotov <ifedotov@suse.com>
Signed-off-by: Igor Fedotov <ifedotov@suse.com>
@ifed01 ifed01 force-pushed the wip-ifed-faster-rm-p1 branch from 9707e0a to ba9b547 Compare December 4, 2020 21:12
@github-actions github-actions bot added the common label Dec 4, 2020
@tchaikov tchaikov merged commit 4ae8abb into ceph:master Dec 5, 2020
@tchaikov
Copy link
Contributor

tchaikov commented Dec 5, 2020

@ifed01 shall we backport this change to, at least, octopus?

@dvanders
Copy link
Contributor

dvanders commented Dec 5, 2020

N too please. (We can take a stab at that if help is needed).

@ifed01 ifed01 deleted the wip-ifed-faster-rm-p1 branch December 7, 2020 17:21
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants