YAOOK Monitoring

kube-state-metrics-values.yaml

customResourceState:
  enabled: true
  config:
    kind: CustomResourceStateMetrics
    spec:
      resources:
        - &resource
          groupVersionKind:
            group: yaook.cloud
            version: "v1"
            kind: "*"
          metricNamePrefix: kube_yaook
          labelsFromPath:
            name: [metadata, name]
            namespace: [metadata, namespace]
          metrics:
            - name: phase_info
              help: "The current phase of a YAOOK resource"
              each:
                type: Info
                info:
                  labelsFromPath:
                    phase: [status, phase]
            - name: status_condition
              help: "The current status conditions of a YAOOK resource"
              each:
                type: Gauge
                gauge:
                  path: [status, conditions]
                  labelsFromPath:
                    condition: ["type"]
                  valueFrom: ["status"]
            - name: created
              help: UNIX creation timestamp
              each:
                type: Gauge
                gauge:
                  path: [metadata, creationTimestamp]
            - name: metadata_generation
              help: Sequence number representing a specific generation of the desired state.
              each:
                type: Gauge
                gauge:
                  path: [metadata, generation]
            - name: status_observed_generation
              help: The generation observed by the YAOOK operator.
              each:
                type: Gauge
                gauge:
                  path: [status, observedGeneration]
            - name: status_updated_generation
              help: Indicates the version of the resource used to generate child resources.
              each:
                type: Gauge
                gauge:
                  path: [status, updatedGeneration]
        - <<: *resource
          groupVersionKind:
            group: compute.yaook.cloud
            version: "v1"
            kind: "*"
        - <<: *resource
          groupVersionKind:
            group: infra.yaook.cloud
            version: "v1"
            kind: "*"
        - <<: *resource
          groupVersionKind:
            group: network.yaook.cloud
            version: "v1"
            kind: "*"
rbac:
  extraRules:
    - apiGroups:
        - yaook.cloud
        - apps.yaook.cloud
        - compute.yaook.cloud
        - infra.yaook.cloud
        - network.yaook.cloud
        - octavia.yaook.cloud
      resources: ['*']
      verbs: ['list', 'watch']

yaook-prometheusrule.yaml

apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata:
  name: yaook-operators
spec:
  groups:
  - name: yaook
    rules:
    - alert: YaookResourceGenerationMismatch
      annotations:
        description: >-
          {{ $labels.customresource_kind }} generation for {{ $labels.namespace }}/{{ $labels.name
          }} does not match, this indicates that the reconciliation has failed but has
          not been rolled back on cluster {{ $labels.cluster }}.
        summary: "{{ $labels.customresource_kind }} generation mismatch due to possible reconciliation error"
      expr: |-
        kube_yaook_status_observed_generation{job="kube-state-metrics", namespace=~".*"}
          !=
        kube_yaook_metadata_generation{job="kube-state-metrics", namespace=~".*"}
      for: 15m
      labels:
        severity: warning
    - alert: YaookResourceReconciliationStuck
      annotations:
        description: Reconciliation of {{ $labels.customresource_kind }} {{ $labels.namespace }}/{{ $labels.name
          }} is not progressing for longer than 15 minutes on cluster {{ $labels.cluster
          }}.
        summary: "{{ $labels.customresource_kind }} reconciliation is not progressing"
      expr: |-
        kube_yaook_status_condition{condition="Converged", job="kube-state-metrics", namespace=~".*"}
        != 1
      for: 15m
      labels:
        severity: warning