Project

General

Profile

Actions
Copy link

Bug #54311

closed

cephadm/monitoring: monitoring stack versions are too old

Added by Paul Cuzner about 2 years ago. Updated over 1 year ago.

Status:
Resolved
Priority:
Normal
Assignee:
Aashish Sharma
Category:
cephadm/monitoring
Target version:
Ceph - v18.0.0
% Done:

100%

Source:
Tags:
Backport:
quincy, pacific
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
45187
Crash signature (v1):
Crash signature (v2):

Description

cephadm is deploying old versions of prometheus, alertmanager, node-exporter, which constrains the features that can be utilised when managing and monitoring the cluster.

Recommend updating to

Prometheus 2.18 -> 2.28
Alertmanager 0.20 -> 0.23
node-exporter 0.18.1 -> 1.3.0

the target versions above are given as a guide only - testing is needed to confirm compatibility

Subtasks 3 (0 open — 3 closed)

Bug #55155: grafana/Makefile: don't push image to dockerResolvedErnesto Puerta

Actions
Bug #55172: monitoring/grafana: fix Grafana versionResolvedErnesto Puerta

Actions
Dashboard - Bug #55195: mgr/dashboard: update grafana piechart and vonage status panel versionsResolvedAashish Sharma

Actions

Related issues 3 (0 open — 3 closed)

Is duplicate of Dashboard - Bug #51957: mgr/dashboard: explore Grafana 7 or 8 migrationDuplicateAashish Sharma

Actions
Copied to Dashboard - Backport #55201: cephadm/monitoring: monitoring stack versions are too oldResolvedAashish SharmaActions
Copied to Dashboard - Backport #55366: cephadm/monitoring: monitoring stack versions are too oldResolvedAashish SharmaActions
Actions
Copy link

Also available in: Atom PDF