Prometheus Adapter

Installs the Prometheus Adapter for the Custom Metrics API. Custom metrics are used in Kubernetes by Horizontal Pod Autoscalers to scale workloads based upon your own metric pulled from an external metrics provider like Prometheus. This chart complements the metrics-server chart that provides resource only metrics.

Prerequisites

Kubernetes 1.14+

Get Helm Repositories Info

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update

See helm repo for command documentation.

Install Helm Chart

helm install [RELEASE_NAME] prometheus-community/prometheus-adapter

See configuration below.

See helm install for command documentation.

Uninstall Helm Chart

helm uninstall [RELEASE_NAME]

This removes all the Kubernetes components associated with the chart and deletes the release.

See helm uninstall for command documentation.

Upgrading Helm Chart

helm upgrade [RELEASE_NAME] [CHART] --install

See helm upgrade for command documentation.

To 4.2.0

Readiness and liveness probes are now fully configurable through values readinessProbe and livenessProbe. The previous values have been kept as defaults.

To 4.0.0

Previously, security context of the container was set directly in the deployment template. This release makes it configurable through the new configuration variable securityContext whilst keeping the previously set values as defaults. Furthermore, previous variable runAsUser is now set in securityContext and is not used any longer. Please, use securityContext.runAsUser instead. In the same security context, seccompProfile has been enabled and set to type RuntimeDefault.

To 3.0.0

Due to a change in deployment labels, the upgrade requires helm upgrade --force in order to re-create the deployment.

Configuration

See Customizing the Chart Before Installing. To see all configurable options with detailed comments, visit the chart's values.yaml, or run these configuration commands:

helm show values prometheus-community/prometheus-adapter

Prometheus Service Endpoint

To use the chart, ensure the prometheus.url and prometheus.port are configured with the correct Prometheus service endpoint. If Prometheus is exposed under HTTPS the host's CA Bundle must be exposed to the container using extraVolumes and extraVolumeMounts.

Adapter Rules

Additionally, the chart comes with a set of default rules out of the box but they may pull in too many metrics or not map them correctly for your needs. Therefore, it is recommended to populate rules.custom with a list of rules (see the config document for the proper format).

Horizontal Pod Autoscaler Metrics

Finally, to configure your Horizontal Pod Autoscaler to use the custom metric, see the custom metrics section of the HPA walkthrough.

The Prometheus Adapter can serve three different metrics APIs:

Custom Metrics

Enabling this option will cause custom metrics to be served at /apis/custom.metrics.k8s.io/v1beta1. Enabled by default when rules.default is true, but can be customized by populating rules.custom:

rules:
  custom:
  - seriesQuery: '{__name__=~"^some_metric_count$"}'
    resources:
      template: <<.Resource>>
    name:
      matches: ""
      as: "my_custom_metric"
    metricsQuery: sum(<<.Series>>{<<.LabelMatchers>>}) by (<<.GroupBy>>)

External Metrics

Enabling this option will cause external metrics to be served at /apis/external.metrics.k8s.io/v1beta1. Can be enabled by populating rules.external:

rules:
  external:
  - seriesQuery: '{__name__=~"^some_metric_count$"}'
    resources:
      template: <<.Resource>>
    name:
      matches: ""
      as: "my_external_metric"
    metricsQuery: sum(<<.Series>>{<<.LabelMatchers>>}) by (<<.GroupBy>>)

Resource Metrics

Enabling this option will cause resource metrics to be served at /apis/metrics.k8s.io/v1beta1. Resource metrics will allow pod CPU and Memory metrics to be used in Horizontal Pod Autoscalers as well as the kubectl top command. Can be enabled by populating rules.resource:

rules:
  resource:
    cpu:
      containerQuery: |
        sum by (<<.GroupBy>>) (
          rate(container_cpu_usage_seconds_total{container!="",<<.LabelMatchers>>}[3m])
        )
      nodeQuery: |
        sum  by (<<.GroupBy>>) (
          rate(node_cpu_seconds_total{mode!="idle",mode!="iowait",mode!="steal",<<.LabelMatchers>>}[3m])
        )
      resources:
        overrides:
          node:
            resource: node
          namespace:
            resource: namespace
          pod:
            resource: pod
      containerLabel: container
    memory:
      containerQuery: |
        sum by (<<.GroupBy>>) (
          avg_over_time(container_memory_working_set_bytes{container!="",<<.LabelMatchers>>}[3m])
        )
      nodeQuery: |
        sum by (<<.GroupBy>>) (
          avg_over_time(node_memory_MemTotal_bytes{<<.LabelMatchers>>}[3m])
          -
          avg_over_time(node_memory_MemAvailable_bytes{<<.LabelMatchers>>}[3m])
        )
      resources:
        overrides:
          node:
            resource: node
          namespace:
            resource: namespace
          pod:
            resource: pod
      containerLabel: container
    window: 3m

NOTE: Setting a value for rules.resource will also deploy the resource metrics API service, providing the same functionality as metrics-server. As such it is not possible to deploy them both in the same cluster.