theshire/kubernetes/apps/observability/victoria-metrics/app/helmrelease.yaml

---
# yaml-language-server: $schema=https://raw.githubusercontent.com/bjw-s/helm-charts/main/charts/other/app-template/schemas/helmrelease-helm-v2beta2.schema.json
apiVersion: helm.toolkit.fluxcd.io/v2beta1
kind: HelmRelease
metadata:
  name: victoria-metrics
spec:
  interval: 30m
  chart:
    spec:
      chart: victoria-metrics-k8s-stack
      version: 0.25.0
      sourceRef:
        kind: HelmRepository
        name: victoria-metrics
        namespace: flux-system
  values:
    fullnameOverride: victoria-metrics

    # VM Operator deployment
    victoria-metrics-operator:
      enabled: true
      operator:
        # disable_prometheus_converter: false # Ensure we keep enabled the converter to sync prom rules to VM rules
        enable_converter_ownership: true # Required to allow VM to remove VM rules it imports if a prometheus rule is deleted

    # Single-binary vm cluster
    vmsingle:
      enabled: true
      spec:
        extraArgs:
          dedup.minScrapeInterval: 30s
          maxLabelsPerTimeseries: "90"
          search.minStalenessInterval: 5m
          vmalert.proxyURL: http://vmalert-victoria-metrics.observability.svc.cluster.local:8080
        retentionPeriod: 1y
        storage:
          storageClassName: "openebs-zfs"
          resources:
            requests:
              storage: "50Gi"
          accessModes:
            - ReadWriteOnce
      ingress:
        enabled: true
        ingressClassName: internal-nginx
        hosts:
          - vm.jahanson.tech

    # VM Alerting (however, this just watches & passes alerts to alertmanager)
    vmalert:
      enabled: true
      spec:
        replicaCount: 1
        resources:
          requests:
            cpu: 50m
            memory: 128Mi
          limits:
            cpu: 150m
            memory: 256Mi
        topologySpreadConstraints:
          - maxSkew: 1
            topologyKey: kubernetes.io/hostname
            whenUnsatisfiable: DoNotSchedule
            labelSelector:
              matchLabels:
                app.kubernetes.io/name: vmalert
        extraArgs:
          external.url: https://vmalert.jahanson.tech
        notifiers:
          - url: http://alertmanager.observability.svc.cluster.local:9093
      ingress:
        enabled: true
        ingressClassName: internal-nginx
        hosts:
          - vmalert.jahanson.tech

    # VM Data scraping
    vmagent:
      enabled: true
      spec:
        replicaCount: 1
        shardCount: 2
        scrapeInterval: 30s
        externalLabels:
          cluster: main
        resources:
          requests:
            cpu: 50m
            memory: 256Mi
          limits:
            cpu: 400m
            memory: 512Mi
        topologySpreadConstraints:
          - maxSkew: 1
            topologyKey: kubernetes.io/hostname
            whenUnsatisfiable: DoNotSchedule
            labelSelector:
              matchLabels:
                app.kubernetes.io/name: vmagent
        additionalScrapeConfigs:
          name: vm-additional-scrape-configs
          key: prometheus-additional.yaml
      ingress:
        enabled: true
        ingressClassName: internal-nginx
        hosts:
          - vmagent.jahanson.tech

    # Extra slack templates
    monzoTemplate:
      enabled: false

    # Scrape configs
    kubelet:
      enabled: true
      spec:
        interval: 30s
        # drop high cardinality label and useless metrics for cadvisor and kubelet
        metricRelabelConfigs:
          - action: labeldrop
            regex: (uid|pod_uid|id)
          - action: labeldrop
            regex: (name)
          - action: drop
            source_labels: [__name__]
            regex: (rest_client_request_duration_seconds_bucket|rest_client_request_duration_seconds_sum|rest_client_request_duration_seconds_count)
          - action: drop
            source_labels: [__name__]
            regex: (container_tasks_state|container_memory_failures_total)
          - action: drop
            source_labels: [__name__]
            regex: (container_blkio_device_usage_total)
          - action: drop
            source_labels: [__name__]
            regex: (prober_probe_duration_seconds_bucket)
        relabelConfigs:
          - action: labelmap
            regex: __meta_kubernetes_node_label_(.+)
          - sourceLabels: [__metrics_path__]
            targetLabel: metrics_path
          - targetLabel: "job"
            replacement: "kubelet"
    # kubeApiServer:
    #   enabled: false
    # kubeControllerManager:
    #   enabled: false
    # kubeProxy:
    #   enabled: false
    # kubeScheduler:
    #   enabled: false
    # kubeEtcd:
    #   enabled: false

    # Enable deployment of kube-state-metrics
    # kube-state-metrics:
    #   enabled: false

    # Enable deployment of prometheus-node-exporter
    # prometheus-node-exporter:
    #   enabled: false

    # Enable deployment of grafana
    # defaultDashboardsEnabled: false
    # grafana:
    #   enabled: false

    # Enable deployment of alertmanager
    # alertmanager:
    #   enabled: false

    # Prepared sets of default rules
    # Adjust to what scraping functions you have enabled
    # i.e. if you dont have kubeapisever setup & enabled, disable
    #   the kubeApiserver rules below
    defaultRules:
      create: true
      rules:
        etcd: false
        general: true
        k8s: true
        kubeApiserver: true
        kubeApiserverAvailability: true
        kubeApiserverBurnrate: true
        kubeApiserverHistogram: true
        kubeApiserverSlos: true
        kubelet: true
        kubePrometheusGeneral: true
        kubePrometheusNodeRecording: true
        kubernetesApps: true
        kubernetesResources: true
        kubernetesStorage: true
        kubernetesSystem: true
        kubeScheduler: false
        kubeStateMetrics: true
        network: true
        node: true
        vmagent: true
        vmsingle: false
        vmhealth: true
        alertmanager: false