Kube-State-Metrics (KSM) Emission

The default Kubecost installation no longer includes a bundled KSM deployment. Instead, Kubecost calculates and emits all required KSM metrics.

KSM metrics emitted by Kubecost

The following table shows all KSM metrics required by and implemented in Kubecost. The below metrics and labels follow conventions of KSMv1, not KSMv2.

Enabling Kubecost-based KSM deployment

If interested in enabling the KSM bundled in Kubecost's Helm chart, perform the below config changes. Re-enabling KSM has the added advantage of high availability. Specifically, if the kubecost-cost-analyzer deployment had downtime, the kube-state-metrics may still be available to emit metrics for Prometheus to scrape.

prometheus:
  kubeStateMetrics:
    enabled: true
  kube-state-metrics:
    disabled: false

Disabling Kubecost's KSM emission

prometheus:
  kubeStateMetrics:
    enabled: false
  kube-state-metrics:
    disabled: true
kubecostMetrics:
  emitKsmV1Metrics: false
  # If you are running KSMv2, you must set the below config as well. More details below.
  emitKsmV1MetricsOnly: true

Disabling individual metrics

It is possible to disable individual metrics emitted by Kubecost if a more fine-grained approach is required. This can be done by setting the related Helm chart parameter:

kubecostProductConfigs:
  ...
  metricsConfigs:
    disabledMetrics:
      - <metric-to-be-disabled>
      - <metric-to-be-disabled>
      etc.

External KSM deployments resulting in duplicated metrics

If your Prometheus deployment is scraping both Kubecost and an external KSM deployment outside of Kubecost, there will be duplicated KSM metrics.

Kubecost itself is resilient to duplicate metrics, but other services or queries could be affected. There are several approaches for handling this problem:

• Remove the external KSM from the cluster. If you do this, only the Kubecost-emitted metrics listed above should be available. However, This could cause other services that depend on KSM metrics to fail.

• Rewrite queries that cannot handle duplicate metrics to include a filter on job=<external-KSM-scrape-job> or to be generally resilient to duplication using query functions like avg_over_time.

• Run a separate Prometheus for Kubecost alone (the default installation behavior of Kubecost) and disable the scraping of Kubecost's metrics in your other Prometheus configurations.

• We support reducing some duplication from Kubecost via config. To reduce the emission of metrics that overlap with metrics provided by KSM v2 you can set the following Helm values (code ref):

  Copy

  kubecostMetrics:
    emitKsmV1MetricsOnly: true
    emitKsmV1Metrics: false

  • The metrics that will still be emitted include:

    • Node metrics (code ref)

      • kube_node_status_capacity

      • kube_node_status_capacity_memory_bytes

      • kube_node_status_capacity_cpu_cores

      • kube_node_status_allocatable

      • kube_node_status_allocatable_memory_bytes

      • kube_node_status_allocatable_cpu_cores

      • kube_node_labels

      • kube_node_status_condition

    • Namespace metrics (code ref)

      • kube_namespace_labels

    • Pod metrics (code ref)

      • kube_pod_labels

      • kube_pod_owner

  • If you are already running KSM v2, and have set the helm value to only emit KSM v1 metrics, you can also disable the Kubecost-based KSM deployment by setting the helm value prometheus.kube-state-metrics.disabled to true.