feat: custom metric sources (#37)

* feat: add custom metric sources * feat: add secrets * fix: remove data from reloader container * style: yamllint * feat: allow custom rules * fix: do not include if empty * fix: empty secrets for default values file * docs: update
agrrh · Dec 14, 2023 · cc26c91 · cc26c91
1 parent 9fdaaf9
commit cc26c91
Show file tree

Hide file tree

Showing 3 changed files with 95 additions and 11 deletions.
diff --git a/deploy/helm/pagetron/templates/prometheus.configmap.yaml b/deploy/helm/pagetron/templates/prometheus.configmap.yaml
@@ -42,7 +42,8 @@ data:
           - targets:
               - blackbox:9115
 
-      # resources
+      # resources: {{ len $.Values.targets }} item(s)
+      {{- if gt (len $.Values.targets) 0 }}
       - job_name: "blackbox"
         metrics_path: /probe
         params:
@@ -60,6 +61,12 @@ data:
             target_label: instance
           - target_label: __address__
             replacement: blackbox:9115
+      {{- end }}
+
+      # custom metrics sources: {{ len $.Values.customMetricSources }} item(s)
+      {{- if gt (len $.Values.customMetricSources) 0 }}
+      {{- toYaml $.Values.customMetricSources | nindent 6 }}
+      {{- end }}
 
   rules_pagetron.yaml: |
     groups:
@@ -68,25 +75,20 @@ data:
         rules:
           - record: pagetron:availability:1m
             expr: |-
-                sum by (instance) (
-                  probe_success == bool 1
-                  and probe_http_status_code >= bool 200
-                  and probe_http_status_code < bool 300
-                  and probe_duration_seconds < bool 2
-                )
+              {{- $.Values.query1m | nindent 14 }}
 
       - name: pagetron_hour
         interval: 60s
         rules:
           - record: pagetron:availability:1h
             expr: |-
-                avg_over_time(pagetron:availability:1m[60m:1m])
+              avg_over_time(pagetron:availability:1m[60m:1m])
 
       - name: pagetron_day
         interval: 30m
         rules:
           - record: pagetron:availability:1d
             expr: |-
-                avg_over_time(pagetron:availability:1h[24h:30m])
+              avg_over_time(pagetron:availability:1h[24h:30m])
 
 {{- end }}
diff --git a/deploy/helm/pagetron/templates/prometheus.deployment.yaml b/deploy/helm/pagetron/templates/prometheus.deployment.yaml
@@ -66,6 +66,12 @@ spec:
             - name: data
               mountPath: /prometheus
 
+            {{- range $secret := $.Values.additionalSecrets }}
+            - name: secret-{{ $secret }}
+              readOnly: true
+              mountPath: "/secrets/{{ $secret }}"
+            {{- end }}
+
           resources:
             {{- toYaml .resources | nindent 12 }}
 
@@ -87,8 +93,6 @@ spec:
           volumeMounts:
             - name: config
               mountPath: /config
-            - name: data
-              mountPath: /prometheus
 
           resources:
             limits:
@@ -108,6 +112,12 @@ spec:
           persistentVolumeClaim:
             claimName: pagetron-prometheus-data
 
+        {{- range $secret := $.Values.additionalSecrets }}
+        - name: secret-{{ $secret }}
+          secret:
+            secretName: {{ $secret }}
+        {{- end }}
+
       {{- with .nodeSelector }}
       nodeSelector:
         {{- toYaml . | nindent 8 }}

diff --git a/deploy/helm/pagetron/values.yaml b/deploy/helm/pagetron/values.yaml
@@ -15,6 +15,78 @@ publicUrl: "https://pagetron.example.org"
 # May be changed to use external Prometheus
 prometheusUrl: "http://prometheus:9090/"
 
+#
+# User may want to change default availability check rule
+#
+# Result MUST be in form of:
+#   - labels:
+#     - "instance" - REQUIRED with target name as value, e.g. "https://example.com"
+#     - other labels allowed
+#   - value: integer, 1 if considered available, otherwise 0
+#
+query1m: |-
+  sum by (instance) (
+    probe_success == bool 1
+    and probe_http_status_code >= bool 200
+    and probe_http_status_code < bool 300
+    and probe_duration_seconds < bool 2
+  )
+
+# State any external metrics here
+# Use scrape_config format:
+#   https://prometheus.io/docs/prometheus/latest/configuration/configuration/#scrape_config
+customMetricSources: []
+  # AWS
+  # Cloudwatch Exporter
+  # https://github.com/prometheus/cloudwatch_exporter
+  #
+  # - job_name: aws-cloudwatch-exporter
+  #   scheme: http
+  #   metrics_path: /metrics
+  #   static_configs:
+  #     - targets:
+  #         - cloudwatch-exporter:9106
+
+  # AWS
+  # Yet Another Cloudwatch Exporter
+  # https://github.com/nerdswords/yet-another-cloudwatch-exporter
+  #
+  # - job_name: aws-yace
+  #   scheme: http
+  #   metrics_path: /metrics
+  #   static_configs:
+  #     - targets:
+  #         - aws-yace:5000
+
+  #
+  # Yandex Cloud
+  # https://cloud.yandex.com/en-ru/docs/monitoring/operations/metric/prometheusExport
+  #
+  # - job_name: yc-monitoring-export
+  #   metrics_path: /monitoring/v2/prometheusMetrics
+  #   scheme: https
+  #   params:
+  #     folderId:
+  #       - my-folder
+  #     service:
+  #       - application-load-balancer
+  #   bearer_token_file: /secrets/yandex-cloud/bearer.key
+  #   static_configs:
+  #     - targets:
+  #         - monitoring.api.cloud.yandex.net
+  #       labels:
+  #         service: application-load-balancer
+  #         folderId: my-folder
+
+# Secret expected to be:
+#  - Created separately
+#  - Located in same namespace
+# These would be mounted to `/secrets` path as separate directories
+# Operator may mention those in custom metric sources list
+additionalSecrets: []
+  # - my-exporter-token
+  # - yandex-cloud
+
 backend:
   replicas: 2
   image: