Merge branch 'main' into fix-opsrecipe-test

CHANGELOG.md

@@ -22,6 +22,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Fix and improve the ops-recipe test script.
 - Fix cabbage alerts for multi-provider wcs.
 - Fix shield alert area labels.
+- Fix `cert-exporter` alerting.
 
 ### Removed
 

helm/prometheus-rules/templates/_helpers.tpl

@@ -45,14 +45,6 @@ phoenix
 {{- end -}}
 {{- end -}}
 
-{{- define "isCertExporterInstalled" -}}
-{{- if has .Values.managementCluster.provider.kind (list "cloud-director" "vsphere" "capa") -}}
-false
-{{- else -}}
-true
-{{- end -}}
-{{- end -}}
-
 {{- define "isBastionBeingMonitored" -}}
 {{ not (eq .Values.managementCluster.provider.flavor "capi") }}
 {{- end -}}

helm/prometheus-rules/templates/shared/alerting-rules/certificate.workload-cluster.rules.yml

@@ -1,4 +1,3 @@
-{{- if eq (include "isCertExporterInstalled" .) "true" }}
 apiVersion: monitoring.coreos.com/v1
 kind: PrometheusRule
 metadata:
@@ -31,7 +30,7 @@ spec:
       annotations:
         description: '{{`Certificate metrics are missing for cluster {{ $labels.cluster_id }}.`}}'
         opsrecipe: absent-metrics
-      expr: max(up{cluster_id!="", cluster_type="workload_cluster"}) by (cluster_id) unless on (cluster_id) count (cert_exporter_not_after{cluster_type="workload_cluster"}) by (cluster_id) > 0
+      expr: max(up{cluster_id!="", cluster_type="workload_cluster"}) by (cluster_id, installation, pipeline, provider) unless on (cluster_id) count (cert_exporter_not_after{cluster_type="workload_cluster"}) by (cluster_id, installation, pipeline, provider) > 0
       for: 30m
       labels:
         area: kaas
@@ -42,4 +41,3 @@ spec:
         severity: page
         team: {{ include "providerTeam" . }}
         topic: security
-{{- end -}}