[stable/fluent-bit] add a servicemonitor resource for Prometheus operator (#14416)

* [stable/fluent-bit] add a servicemonitor resource for Prometheus operator

Signed-off-by: Sébastien Prud'homme <sebastien.prudhomme@gmail.com>

* [stable/fluent-bit] check API availability

Signed-off-by: Sébastien Prud'homme <sebastien.prudhomme@gmail.com>

stable/fluent-bit/Chart.yaml

@@ -1,6 +1,6 @@
 apiVersion: v1
 name: fluent-bit
-version: 2.3.0
+version: 2.4.0
 appVersion: 1.2.1
 description: Fast and Lightweight Log/Data Forwarder for Linux, BSD and OSX
 keywords:

stable/fluent-bit/README.md

@@ -130,6 +130,11 @@ The following table lists the configurable parameters of the Fluent-Bit chart an
 | `metrics.service.labels`           | Additional labels for the fluent-bit metrics service definition, specified as a map.                                                    | None                                              |
 | `metrics.service.port`             | Port on where metrics should be exposed    | `2020`                                            |
 | `metrics.service.type`             | Service type for metrics                   | `ClusterIP`                                       |
+| `metrics.serviceMonitor.enabled`          | Set this to `true` to create ServiceMonitor for Prometheus operator                   | `false` |
+| `metrics.serviceMonitor.additionalLabels` | Additional labels that can be used so ServiceMonitor will be discovered by Prometheus | `{}`    |
+| `metrics.serviceMonitor.namespace`        | Optional namespace in which to create ServiceMonitor                                  | `nil`   |
+| `metrics.serviceMonitor.interval`         | Scrape interval. If not set, the Prometheus default scrape interval is used           | `nil`   |
+| `metrics.serviceMonitor.scrapeTimeout`    | Scrape timeout. If not set, the Prometheus default scrape timeout is used             | `nil`   |
 | `trackOffsets`                     | Specify whether to track the file offsets for tailing docker logs. This allows fluent-bit to pick up where it left after pod restarts but requires access to a `hostPath` | `false` |
 | `testFramework.image`              | `test-framework` image repository.         | `dduportal/bats`                                  |
 | `testFramework.tag`                | `test-framework` image tag.                | `0.4.0`                                           |

stable/fluent-bit/templates/servicemonitor.yaml

@@ -0,0 +1,34 @@
+{{- if and .Values.metrics.enabled .Values.metrics.serviceMonitor.enabled (.Capabilities.APIVersions.Has "monitoring.coreos.com/v1") }}
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+  name: {{ include "fluent-bit.fullname" . }}
+  {{- if .Values.metrics.serviceMonitor.namespace }}
+  namespace: {{ .Values.metrics.serviceMonitor.namespace }}
+  {{- end }}
+  labels:
+    app: {{ template "fluent-bit.name" . }}
+    chart: {{ .Chart.Name }}-{{ .Chart.Version }}
+    heritage: {{ .Release.Service }}
+    release: {{ .Release.Name }}
+    {{- if .Values.metrics.serviceMonitor.additionalLabels }}
+{{ toYaml .Values.metrics.serviceMonitor.additionalLabels | indent 4 }}
+    {{- end }}
+spec:
+  endpoints:
+    - port: metrics
+      path: /api/v1/metrics/prometheus
+      {{- if .Values.metrics.serviceMonitor.interval }}
+      interval: {{ .Values.metrics.serviceMonitor.interval }}
+      {{- end }}
+      {{- if .Values.metrics.serviceMonitor.scrapeTimeout }}
+      scrapeTimeout: {{ .Values.metrics.serviceMonitor.scrapeTimeout }}
+      {{- end }}
+  namespaceSelector:
+    matchNames:
+      - {{ .Release.Namespace }}
+  selector:
+    matchLabels:
+      app: {{ include "fluent-bit.name" . }}
+      release: {{ .Release.Name }}
+{{- end }}

stable/fluent-bit/values.yaml

@@ -28,6 +28,12 @@ metrics:
     # prometheus.io/scrape: "true"
     port: 2020
     type: ClusterIP
+  serviceMonitor:
+    enabled: false
+    additionalLabels: {}
+    # namespace: monitoring
+    # interval: 30s
+    # scrapeTimeout: 10s
 
 # When enabled, fluent-bit will keep track of tailing offsets across pod restarts.
 trackOffsets: false