Project

General

Profile

Bug #42961

osd: increase priority in certain OSD perf counters

Added by Ernesto Puerta over 4 years ago. Updated about 4 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
Category:
Administration/Usability
Target version:
% Done:

0%

Source:
Tags:
Backport:
nautilus
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
BlueStore, OSD
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

There are reports from users missing stats in dashboard/prometheus mgr modules about the following perf counters:

        "bluestore_compressed",
        "bluestore_compressed_allocated",
        "bluestore_compressed_original",

They all are tagged as priority 0 (PRIO_DEBUGONLY), and ceph-mgr filters (by default) out any perf counter below 5 (PRIO_USEFUL). There are 2 possible approaches here:
  • Ad-hoc fix by increasing the priority for these specific perf counters from 0 to 5.
  • Longer term one: by providing a mgr option to cherry-pick perf counters of lower than default priority and export them.

Related issues

Related to mgr - Feature #46212: mgr/prometheus: allow users to export metrics of lower priorities In Progress
Copied to RADOS - Backport #43245: nautilus: osd: increase priority in certain OSD perf counters Resolved

History

#1 Updated by Neha Ojha over 4 years ago

  • Assignee set to Neha Ojha

#2 Updated by Neha Ojha over 4 years ago

  • Status changed from New to Fix Under Review
  • Pull request ID set to 31959

Ernesto, while we are at it, are there any other specific stats that you've gotten requests for?

#3 Updated by Neha Ojha over 4 years ago

  • Backport set to nautilus

#4 Updated by Ernesto Puerta over 4 years ago

Neha Ojha wrote:

Ernesto, while we are at it, are there any other specific stats that you've gotten requests for?

Neha, not that I know of. This comes from a downstream BZ (https://bugzilla.redhat.com/show_bug.cgi?id=1679924) and a customer ticket. Perhaps, we could move the discussion to users & customer facing experts, as they would probably know better what stats they missed in Prometheus/Dashboard. I'll send a mail to ceph-users on this regard. Nevertheless, I think we can go with this as it is.

Thanks a lot for taking care of it!

#5 Updated by Neha Ojha over 4 years ago

  • Status changed from Fix Under Review to Pending Backport

#6 Updated by Nathan Cutler over 4 years ago

  • Copied to Backport #43245: nautilus: osd: increase priority in certain OSD perf counters added

#7 Updated by Nathan Cutler about 4 years ago

  • Status changed from Pending Backport to Resolved

While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are in status "Resolved" or "Rejected".

#8 Updated by Ernesto Puerta almost 4 years ago

  • Related to Feature #46212: mgr/prometheus: allow users to export metrics of lower priorities added

Also available in: Atom PDF