Power Monitoring

Power monitoring for Red Hat OpenShift enables you to monitor the power usage of workloads and identify the most power-consuming namespaces running in an OpenShift Container Platform cluster with key power consumption metrics, such as CPU or DRAM, measured at container level.

This release of power monitoring and the Power Monitoring Operator provides more accurate data, includes new dashboards, and removes some features and functionality.

This release of power monitoring and the Power Monitoring Operator is supported on:

This release includes the following version updates:

This release includes the following version updates:

The following advisory is available for power monitoring 0.3:

This release includes the following version updates:

The following advisory is available for power monitoring 0.2:

This release introduces a Technology Preview version of power monitoring for Red Hat OpenShift. The following advisory is available for power monitoring 0.1:

You can use power monitoring for Red Hat OpenShift to monitor the power usage and identify power-consuming containers running in an OpenShift Container Platform cluster. Power monitoring collects and exports energy-related system statistics from various components, such as CPU and DRAM. It provides estimates and granular power consumption data for Kubernetes pods and namespaces, and reads the power consumption of nodes.

Power monitoring is made up of the following major components:

Kepler is the key component of power monitoring that collects real-time CPU power consumption data from a node through the RAPL Subsystem. By understanding the total power consumption of the node and calculating the percent of CPU time each process is using, it is able to estimate the power consumption at a per process and container level.

Starting with version 0.4, Power Monitoring Operator for Red Hat OpenShift is FIPS compliant. When deployed on an OpenShift Container Platform cluster in FIPS mode, it uses Red Hat Enterprise Linux (RHEL) cryptographic libraries validated by National Institute of Standards and Technology (NIST).

For details on the NIST validation program, see Cryptographic module validation program. For the latest NIST status of RHEL cryptographic libraries, see Compliance activities and government standards.

To enable FIPS mode, you must install Power Monitoring Operator for Red Hat OpenShift on an OpenShift Container Platform cluster. For more information, see "Do you need extra security for your cluster?".

As a cluster administrator, you can install the Power Monitoring Operator from OperatorHub by using the OpenShift Container Platform web console.

You can deploy Kepler by creating an instance of the PowerMonitor custom resource (CR) using the Power Monitoring Operator.

You can configure Kepler with the spec field of the PowerMonitor resource.

The following is the list of configuration options:

apiVersion: v1alpha1
kind: PowerMonitor
metadata:
  labels:
    app.kubernetes.io/name: powermonitor
    app.kubernetes.io/instance: powermonitor
    app.kubernetes.io/part-of: kepler-operator
  name: power-monitor
spec:
  kepler:
    deployment:
      nodeSelector:
        kubernetes.io/os: linux

      tolerations:
        - key: key1
          operator: Equal
          value: value1
          effect: NoSchedule

      security:
        mode: rbac
        allowedSANames:
          - openshift-user-workload-monitoring:prometheus-user-workload

    config:
      logLevel: info
      metricLevels: [node, pod, vm]
      staleness: 1s
      sampleRate: 10s
      maxTerminated: 1000

apiVersion: v1alpha1
kind: PowerMonitor
metadata:
  labels:
    app.kubernetes.io/name: powermonitor
    app.kubernetes.io/instance: powermonitor
    app.kubernetes.io/part-of: kepler-operator
  name: power-monitor
spec:
  kepler:
    deployment:
      nodeSelector:
        kubernetes.io/os: linux

      tolerations:
        - key: key1
          operator: Equal
          value: value1
          effect: NoSchedule

      security:
        mode: rbac
        allowedSANames:
          - openshift-user-workload-monitoring:prometheus-user-workload

    config:
      logLevel: info
      metricLevels: [node, pod, vm]
      staleness: 1s
      sampleRate: 10s
      maxTerminated: 1000

You can monitor the state of the Kepler exporter with the status field of the PowerMonitor resource.

The status field includes information, such as the following:

This provides you with valuable insights into the changes made through the spec field.

apiVersion: kepler.system.sustainable.computing.io/v1alpha1
kind: PowerMonitor
metadata:
  name: power-monitor
status:
   conditions: 
     - lastTransitionTime: '2024-01-11T11:07:39Z'
       message: Reconcile succeeded
       observedGeneration: 1
       reason: ReconcileSuccess
       status: 'True'
       type: Reconciled
     - lastTransitionTime: '2024-01-11T11:07:39Z'
       message: >-
         power-monitor daemonset "openshift-power-monitoring/power-monitor" is deployed to all nodes and
         available; ready 2/2
       observedGeneration: 1
       reason: DaemonSetReady
       status: 'True'
       type: Available
   currentNumberScheduled: 2 
   desiredNumberScheduled: 2 

apiVersion: kepler.system.sustainable.computing.io/v1alpha1
kind: PowerMonitor
metadata:
  name: power-monitor
status:
   conditions: 

     - lastTransitionTime: '2024-01-11T11:07:39Z'
       message: Reconcile succeeded
       observedGeneration: 1
       reason: ReconcileSuccess
       status: 'True'
       type: Reconciled
     - lastTransitionTime: '2024-01-11T11:07:39Z'
       message: >-
         power-monitor daemonset "openshift-power-monitoring/power-monitor" is deployed to all nodes and
         available; ready 2/2
       observedGeneration: 1
       reason: DaemonSetReady
       status: 'True'
       type: Available
   currentNumberScheduled: 2 

   desiredNumberScheduled: 2 

There are two types of power monitoring dashboards. Both provide different levels of details around power consumption metrics for a single cluster:

You can access power monitoring dashboards of the OpenShift Container Platform web console.

You can access power monitoring dashboards from OpenShift Container Platform web console.

You can delete Kepler by removing the Kepler instance of the Kepler custom resource definition (CRD) from the OpenShift Container Platform web console.

You can delete the PowerMonitor custom resource (CR) by removing the power-monitor instance of the PowerMonitor CR from the OpenShift Container Platform web console.

If you installed the Power Monitoring Operator by using OperatorHub, you can uninstall it from the OpenShift Container Platform web console.

Kepler’s power attribution system provides practical, proportional distribution of hardware energy consumption to individual workloads. While CPU-time-based attribution has inherent limitations due to modern CPU complexity, it offers a good balance between accuracy, simplicity, and performance overhead for most monitoring and optimization use cases.

For more information about power attribution, see Kepler Power Attribution Guide.

PowerMonitor

PowerMonitor is the schema for the PowerMonitor API.