Fix or replace dashboards using Angular components

roles/certmanager/tasks/main.yml

@@ -36,22 +36,28 @@
     wait: yes
     wait_timeout: "{{ certmanager_wait_timeout }}"
 
-- name: Install Grafana dashboard for cert-manager metrics
-  command: kubectl apply -f -
-  args:
-    stdin: "{{ certmanager_dashboard_definition | to_nice_yaml }}"
-  vars:
-    certmanager_dashboard_definition:
-      apiVersion: v1
-      kind: ConfigMap
-      metadata:
-        name: cert-manager-grafana-dashboard
-        namespace: "{{ certmanager_release_namespace }}"
-        labels:
-          grafana_dashboard: "1"
-      data:
-        certmanager_dashboard.json: |-
-          {{ lookup("file", "grafana_dashboard.json") | from_json | to_nice_json }}
+- block:
+    - name: Install Grafana dashboard for cert-manager metrics
+      command: kubectl apply -f -
+      args:
+        stdin: "{{ certmanager_dashboard_definition | to_nice_yaml }}"
+      vars:
+        certmanager_dashboard_definition:
+          apiVersion: v1
+          kind: ConfigMap
+          metadata:
+            name: cert-manager-grafana-dashboard
+            namespace: "{{ certmanager_release_namespace }}"
+            labels:
+              grafana_dashboard: "1"
+          data:
+            certmanager_dashboard.json: |-
+              {{ lookup("file", "grafana_dashboard.json") | from_json | to_nice_json }}
+
+    - name: Configure custom alerting rules for cert-manager
+      command: kubectl apply -f -
+      args:
+        stdin: "{{ lookup('template', 'prometheusrule.yaml.j2') }}"
   when: certmanager_monitoring_enabled
 
 - block:

roles/certmanager/templates/prometheusrule.yaml.j2

@@ -0,0 +1,53 @@
+apiVersion: monitoring.coreos.com/v1
+kind: PrometheusRule
+metadata:
+  name: cert-manager-alerting-rules
+  namespace:  "{{ certmanager_release_namespace }}"
+  labels:
+    release: kube-prometheus-stack
+{% raw %}
+spec:
+  groups:
+    - name: cert-manager.rules
+      rules:
+        - alert: CertManagerAbsent
+          annotations:
+            description: >-
+              New certificates will not be able to be minted, and existing ones can't
+              be renewed until cert-manager is back.
+            runbook_url: https://github.com/imusmanmalik/cert-manager-mixin/blob/main/RUNBOOK.md#certmanagerabsent
+            summary: Cert Manager has disappeared from Prometheus service discovery.
+          expr: absent(up{job="cert-manager"})
+          for: 10m
+          labels:
+            severity: critical
+
+        - alert: CertManagerCertExpirySoon
+          annotations:
+            description: >-
+              The domain that this cert covers will be unavailable after {{ $value | humanizeDuration }}.
+              Clients using endpoints that this cert protects will start to fail in {{ $value | humanizeDuration }}.
+            runbook_url: https://github.com/imusmanmalik/cert-manager-mixin/blob/main/RUNBOOK.md#certmanagercertexpirysoon
+            summary: The cert `{{ $labels.name }}` is {{ $value | humanizeDuration }} from expiry.
+          expr: |
+            avg by (exported_namespace, namespace, name) (
+              certmanager_certificate_expiration_timestamp_seconds - time()
+            ) < (30 * 24 * 3600)
+          for: 1h
+          labels:
+            severity: warning
+
+        - alert: CertManagerHittingRateLimits
+          annotations:
+            description: >-
+              Depending on the rate limit, cert-manager may be unable to generate certificates for up to a week.
+            runbook_url: https://github.com/imusmanmalik/cert-manager-mixin/blob/main/RUNBOOK.md#certmanagerhittingratelimits
+            summary: Cert manager hitting LetsEncrypt rate limits.
+          expr: |
+            sum by (host) (
+              rate(certmanager_http_acme_client_request_count{status="429"}[5m])
+            ) > 0
+          for: 5m
+          labels:
+            severity: critical
+{% endraw %}