From 5ea26c7421413b543c485c043d8a999f090830ec Mon Sep 17 00:00:00 2001
From: Jose Armesto <github@armesto.net>
Date: Wed, 20 Nov 2024 14:51:37 +0100
Subject: [PATCH] Use generic alert for all providers

---
 CHANGELOG.md                                  |  3 +-
 .../azure-cloud-components.rules.yml          | 32 -------------------
 ...ml => cloud-provider-controller.rules.yml} | 23 +++++++++----
 3 files changed, 18 insertions(+), 40 deletions(-)
 delete mode 100644 helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/azure-cloud-components.rules.yml
 rename helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/{aws-cloud-components.rules.yml => cloud-provider-controller.rules.yml} (54%)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 1a96c195..ec246a20 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,8 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Added
 
-- Add `aws-cloud-components.rules` to monitor the AWS cloud-controller and the ebs-csi-driver.
-- Add `azure-cloud-components.rules` to monitor the Azure cloud-controller and the azure csi drivers.
+- Add `cloud-provider-controller.rules` to monitor the cloud-provider-controller components across providers.
 - Add alert to monitor the HelmRelease for vertical-pod-autoscaler-crd app.
 
 ## [4.26.1] - 2024-11-19
diff --git a/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/azure-cloud-components.rules.yml b/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/azure-cloud-components.rules.yml
deleted file mode 100644
index e17f0029..00000000
--- a/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/azure-cloud-components.rules.yml
+++ /dev/null
@@ -1,32 +0,0 @@
-{{- if eq .Values.managementCluster.provider.kind "capz" }}
-# This rule applies to capa management clusters only
-apiVersion: monitoring.coreos.com/v1
-kind: PrometheusRule
-metadata:
-  creationTimestamp: null
-  labels:
-    {{- include "labels.common" . | nindent 4 }}
-  name: azure-cloud-components.rules
-  namespace: {{ .Values.namespace  }}
-spec:
-  groups:
-  - name: azure-cloud-components
-    rules:
-    - alert: FluxHelmReleaseFailed
-      annotations:
-        description: |-
-          {{`Flux HelmRelease {{ $labels.name }} in ns {{ $labels.exported_namespace }} on {{ $labels.installation }}/{{ $labels.cluster_id }} is stuck in Failed state.`}}
-        opsrecipe: fluxcd-failing-helmrelease/
-      expr: gotk_reconcile_condition{type="Ready", status="False", kind="HelmRelease", cluster_type="management_cluster", exported_namespace!="flux-giantswarm", name=~".*(azure-cloud-controller-manager|azure-cloud-node-manager|azuredisk-csi-driver|azurefile-csi-driver)"} > 0
-      for: 20m
-      labels:
-        area: kaas
-        cancel_if_outside_working_hours: "true"
-        cancel_if_kube_state_metrics_down: "true"
-        cancel_if_monitoring_agent_down: "true"
-        severity: page
-        team: phoenix
-        topic: managementcluster
-        namespace: |-
-          {{`{{ $labels.exported_namespace }}`}}
-{{- end }}
diff --git a/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/aws-cloud-components.rules.yml b/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/cloud-provider-controller.rules.yml
similarity index 54%
rename from helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/aws-cloud-components.rules.yml
rename to helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/cloud-provider-controller.rules.yml
index c77ea62d..5a218b53 100644
--- a/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/aws-cloud-components.rules.yml
+++ b/helm/prometheus-rules/templates/kaas/phoenix/alerting-rules/cloud-provider-controller.rules.yml
@@ -1,23 +1,34 @@
-{{- if eq .Values.managementCluster.provider.kind "capa" }}
-# This rule applies to capa management clusters only
+{{- if eq .Values.managementCluster.provider.flavor "capi" }}
+# This rule applies to CAPI management clusters only
+{{- define "cloudProviderControllerComponents" -}}
+- aws-ebs-csi-driver
+- cloud-provider-aws
+- azure-cloud-controller-manager
+- azure-cloud-node-manager
+- azuredisk-csi-driver
+- azurefile-csi-driver
+- cloud-provider-vsphere
+- cloud-provider-cloud-director
+{{- end }}
 apiVersion: monitoring.coreos.com/v1
 kind: PrometheusRule
 metadata:
   creationTimestamp: null
   labels:
     {{- include "labels.common" . | nindent 4 }}
-  name: aws-cloud-components.rules
+  name: cloud-provider-controller.rules
   namespace: {{ .Values.namespace  }}
 spec:
   groups:
-  - name: aws-cloud-components
+  - name: cloud-provider-controller
     rules:
     - alert: FluxHelmReleaseFailed
       annotations:
         description: |-
           {{`Flux HelmRelease {{ $labels.name }} in ns {{ $labels.exported_namespace }} on {{ $labels.installation }}/{{ $labels.cluster_id }} is stuck in Failed state.`}}
         opsrecipe: fluxcd-failing-helmrelease/
-      expr: gotk_reconcile_condition{type="Ready", status="False", kind="HelmRelease", cluster_type="management_cluster", exported_namespace!="flux-giantswarm", name=~".*(aws-ebs-csi-driver|cloud-provider-aws)"} > 0
+      # Here we take the list of components from the cloudProviderControllerComponents template function and transform it into a |-separated string, which is suitable for the PromQL query
+      expr: gotk_reconcile_condition{type="Ready", status="False", kind="HelmRelease", cluster_type="management_cluster", exported_namespace!="flux-giantswarm", name=~".*{{ (include "cloudProviderControllerComponents" . | fromYaml | join "\\|") }}"} > 0
       for: 20m
       labels:
         area: kaas
@@ -25,7 +36,7 @@ spec:
         cancel_if_kube_state_metrics_down: "true"
         cancel_if_monitoring_agent_down: "true"
         severity: page
-        team: phoenix
+        team: {{ include "providerTeam" . }}
         topic: managementcluster
         namespace: |-
           {{`{{ $labels.exported_namespace }}`}}