Added ServiceMonitor/Prometheus Operator support as like in the original fluent-bit package

[szwed]

We are using Prometheus Operator for monitoring all the middleware components (including fluent-bit/humio-fluentbit) therefore we need built in support for ServiceMonitor CRDs as like it was done in the original chart:
https://github.com/helm/charts/blob/master/stable/fluent-bit/templates/servicemonitor.yaml

This is community wide standard and most of the community driven stable charts provide the functionality.

Changes:

If both the attributes are set to true:

metrics:
  enabled: true
    serviceMonitor:
      enabled: true

The following new objects are being created:

# Source: humio-fluentbit/templates/fluent-bit-servicemonitor.yaml
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  name: RELEASE-NAME-humio-fluentbit
  labels:
    app: humio-fluentbit
    chart: humio-fluentbit-0.1.0
    heritage: Helm
    release: RELEASE-NAME
spec:
  endpoints:
    - port: metrics
      path: /api/v1/metrics/prometheus
  namespaceSelector:
    matchNames:
      - default
  selector:
    matchLabels:
      app: humio-fluentbit
      release: RELEASE-NAME

# Source: humio-fluentbit/templates/fluent-bit-service.yaml
apiVersion: v1
kind: Service
metadata:
  name: RELEASE-NAME-humio-fluentbit-metrics
  labels:
    app: humio-fluentbit
    chart: humio-fluentbit-0.1.0
    heritage: Helm
    release: RELEASE-NAME
spec:
  type: ClusterIP
  sessionAffinity: None
  ports:
  - port: 2020
    targetPort: metrics
    name: metrics
  selector:
    app: humio-fluentbit
    release: RELEASE-NAME

and fluent-bit config gets extended with:

  fluent-bit-service.conf:  |-
    [SERVICE]
        Flush        1
        Daemon       Off
        Log_Level    info
        Parsers_File parsers.conf
        HTTP_Server  On                    # new !!!!
        HTTP_Listen  0.0.0.0              # new !!!!
        HTTP_Port    2020                  # new !!!!