[piraeus-operator] Fix LINSTOR satellite alert labels and scrape flapping false positives

Signed-off-by: sasha-sup <alexsup4ik@gmail.com>
(cherry picked from commit a562d9cb80)
This commit is contained in:
sasha-sup 2026-03-25 15:42:07 +03:00 committed by github-actions[bot]
parent e77ee47e15
commit f7161c2af8

View file

@ -17,9 +17,12 @@ spec:
- alert: linstorSatelliteErrorRate
annotations:
description: |
LINSTOR Satellite "{{ $labels.name }}" reports {{ $value }} errors in the last 15 minutes.
Use "linstor error-reports list --nodes {{ $labels.name }} --since 15minutes" to see them.
expr: increase(linstor_error_reports_count{module="SATELLITE"}[15m]) > 0
LINSTOR Satellite "{{ $labels.hostname }}" reports {{ $value }} errors in the last 15 minutes.
Use "linstor error-reports list --nodes {{ $labels.hostname }}" to inspect the reports.
expr: |
increase(linstor_error_reports_count{module="SATELLITE"}[15m]) > 0
and on(instance, job)
min_over_time(up{job="linstor-controller"}[15m]) == 1
labels:
severity: warning
- alert: linstorControllerErrorRate
@ -33,7 +36,7 @@ spec:
- alert: linstorSatelliteNotOnline
annotations:
description: |
LINSTOR Satellite "{{ $labels.name }}" is not ONLINE.
LINSTOR Satellite "{{ $labels.hostname }}" is not ONLINE.
Check that the Satellite is running and reachable from the LINSTOR Controller.
expr: linstor_node_state{nodetype="SATELLITE"} != 2
labels: