pkg/manifests: Deploy metrics-server through operator

Signed-off-by: Jayapriya Pai <janantha@redhat.com>

Documentation/api.md

@@ -25,7 +25,7 @@ Configuring Cluster Monitoring is optional. If the config does not exist or is e
 * [DedicatedServiceMonitors](#dedicatedservicemonitors)
 * [K8sPrometheusAdapter](#k8sprometheusadapter)
 * [KubeStateMetricsConfig](#kubestatemetricsconfig)
-* [MetricsServer](#metricsserver)
+* [MetricsServerConfig](#metricsserverconfig)
 * [MonitoringPluginConfig](#monitoringpluginconfig)
 * [NodeExporterCollectorBuddyInfoConfig](#nodeexportercollectorbuddyinfoconfig)
 * [NodeExporterCollectorConfig](#nodeexportercollectorconfig)
@@ -125,8 +125,7 @@ The `ClusterMonitoringConfiguration` resource defines settings that customize th
 | -------- | ---- | ----------- |
 | alertmanagerMain | *[AlertmanagerMainConfig](#alertmanagermainconfig) | `AlertmanagerMainConfig` defines settings for the Alertmanager component in the `openshift-monitoring` namespace. |
 | enableUserWorkload | *bool | `UserWorkloadEnabled` is a Boolean flag that enables monitoring for user-defined projects. |
-| enableMetricsServer | *bool | `MetricsServerEnabled` is a Boolean flag that enables metrics-server. |
-| metricsServer | *[MetricsServer](#metricsserver) | `MetricsServer` defines settings for the MetricsServer component. |
+| metricsServer | *[MetricsServerConfig](#metricsserverconfig) | `MetricsServer` defines settings for the MetricsServer component. |
 | k8sPrometheusAdapter | *[K8sPrometheusAdapter](#k8sprometheusadapter) | `K8sPrometheusAdapter` defines settings for the Prometheus Adapter component. |
 | kubeStateMetrics | *[KubeStateMetricsConfig](#kubestatemetricsconfig) | `KubeStateMetricsConfig` defines settings for the `kube-state-metrics` agent. |
 | prometheusK8s | *[PrometheusK8sConfig](#prometheusk8sconfig) | `PrometheusK8sConfig` defines settings for the Prometheus component. |
@@ -188,20 +187,21 @@ The `KubeStateMetricsConfig` resource defines settings for the `kube-state-metri
 
 [Back to TOC](#table-of-contents)
 
-## MetricsServer
+## MetricsServerConfig
 
 #### Description
 
-The `MetricsServer` resource defines settings for the MetricsServer component.
+The `MetricsServerConfig` resource defines settings for the MetricsServer component.
 
 
 <em>appears in: [ClusterMonitoringConfiguration](#clustermonitoringconfiguration)</em>
 
 | Property | Type | Description |
 | -------- | ---- | ----------- |
+| enabled | *bool | `Enabled` is a Boolean flag that enables or disables metrics-server. |
 | nodeSelector | map[string]string | Defines the nodes on which the pods are scheduled. |
 | tolerations | [][v1.Toleration](https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.26/#toleration-v1-core) | Defines tolerations for the pods. |
-| resources | *[v1.ResourceRequirements](https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.26/#resourcerequirements-v1-core) | Defines resource requests and limits for the Prometheus container. |
+| resources | *[v1.ResourceRequirements](https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.26/#resourcerequirements-v1-core) | Defines resource requests and limits for the Metrics Server container. |
 
 [Back to TOC](#table-of-contents)
 

Documentation/openshiftdocs/index.adoc

@@ -45,7 +45,7 @@ The configuration file itself is always defined under the `config.yaml` key in t
 * link:modules/dedicatedservicemonitors.adoc[DedicatedServiceMonitors]
 * link:modules/k8sprometheusadapter.adoc[K8sPrometheusAdapter]
 * link:modules/kubestatemetricsconfig.adoc[KubeStateMetricsConfig]
-* link:modules/metricsserver.adoc[MetricsServer]
+* link:modules/metricsserverconfig.adoc[MetricsServerConfig]
 * link:modules/monitoringpluginconfig.adoc[MonitoringPluginConfig]
 * link:modules/nodeexportercollectorbuddyinfoconfig.adoc[NodeExporterCollectorBuddyInfoConfig]
 * link:modules/nodeexportercollectorconfig.adoc[NodeExporterCollectorConfig]

Documentation/openshiftdocs/modules/clustermonitoringconfiguration.adoc

@@ -19,9 +19,7 @@ The `ClusterMonitoringConfiguration` resource defines settings that customize th
 
 |enableUserWorkload|*bool|`UserWorkloadEnabled` is a Boolean flag that enables monitoring for user-defined projects.
 
-|enableMetricsServer|*bool|`MetricsServerEnabled` is a Boolean flag that enables metrics-server.
-
-|metricsServer|*link:metricsserver.adoc[MetricsServer]|`MetricsServer` defines settings for the MetricsServer component.
+|metricsServer|*link:metricsserverconfig.adoc[MetricsServerConfig]|`MetricsServer` defines settings for the MetricsServer component.
 
 |k8sPrometheusAdapter|*link:k8sprometheusadapter.adoc[K8sPrometheusAdapter]|`K8sPrometheusAdapter` defines settings for the Prometheus Adapter component.
 

Documentation/openshiftdocs/modules/metricsserver.adoc

@@ -18,11 +18,13 @@ Appears in: link:clustermonitoringconfiguration.adoc[ClusterMonitoringConfigurat
 [options="header"]
 |===
 | Property | Type | Description 
+|enableMetricsServer|*bool|`EnableMetricsServer` is a Boolean flag that enables metrics-server.
+
 |nodeSelector|map[string]string|Defines the nodes on which the pods are scheduled.
 
 |tolerations|[]v1.Toleration|Defines tolerations for the pods.
 
-|resources|*v1.ResourceRequirements|Defines resource requests and limits for the Prometheus container.
+|resources|*v1.ResourceRequirements|Defines resource requests and limits for the Metrics Server container.
 
 |===
 

Documentation/openshiftdocs/modules/metricsserverconfig.adoc

@@ -0,0 +1,31 @@
+// DO NOT EDIT THE CONTENT IN THIS FILE. It is automatically generated from the 
+	// source code for the Cluster Monitoring Operator. Any changes made to this 
+	// file will be overwritten when the content is re-generated. If you wish to 
+	// make edits, read the docgen utility instructions in the source code for the 
+	// CMO.
+	:_content-type: ASSEMBLY
+
+== MetricsServerConfig
+
+=== Description
+
+The `MetricsServerConfig` resource defines settings for the MetricsServer component.
+
+
+
+Appears in: link:clustermonitoringconfiguration.adoc[ClusterMonitoringConfiguration]
+
+[options="header"]
+|===
+| Property | Type | Description 
+|enabled|*bool|`Enabled` is a Boolean flag that enables or disables metrics-server.
+
+|nodeSelector|map[string]string|Defines the nodes on which the pods are scheduled.
+
+|tolerations|[]v1.Toleration|Defines tolerations for the pods.
+
+|resources|*v1.ResourceRequirements|Defines resource requests and limits for the Metrics Server container.
+
+|===
+
+link:../index.adoc[Back to TOC]

manifests/image-references

@@ -70,3 +70,7 @@ spec:
     from:
       kind: DockerImage
       name: quay.io/openshift/origin-monitoring-plugin:latest
+  - name: metrics-server
+    from:
+      kind: DockerImage
+      name: registry.k8s.io/metrics-server/metrics-server:v0.6.3

pkg/client/client.go

@@ -915,6 +915,15 @@ func (c *Client) DeleteSecret(ctx context.Context, s *v1.Secret) error {
 	return err
 }
 
+func (c *Client) DeleteAPIService(ctx context.Context, apiService *apiregistrationv1.APIService) error {
+	err := c.aggclient.ApiregistrationV1().APIServices().Delete(ctx, apiService.GetName(), metav1.DeleteOptions{})
+	if apierrors.IsNotFound(err) {
+		return nil
+	}
+
+	return err
+}
+
 func (c *Client) WaitForPrometheus(ctx context.Context, p *monv1.Prometheus) error {
 	var lastErr error
 	if err := wait.Poll(time.Second*10, time.Minute*5, func() (bool, error) {

pkg/manifests/config.go

@@ -125,6 +125,7 @@ func (c Config) GetThanosRulerAlertmanagerConfigs() []AdditionalAlertmanagerConf
 }
 
 type Images struct {
+	MetricsServer                      string
 	K8sPrometheusAdapter               string
 	PromLabelProxy                     string
 	PrometheusOperatorAdmissionWebhook string
@@ -223,8 +224,8 @@ func NewConfig(content io.Reader, tp bool) (*Config, error) {
 		return nil, errors.Wrap(ErrConfigValidation, fmt.Sprintf(`%q is not supported, supported collection profiles are: %q`, c.ClusterMonitoringConfiguration.PrometheusK8sConfig.CollectionProfile, SupportedCollectionProfiles.String()))
 	}
 
-	if c.ClusterMonitoringConfiguration.MetricsServerEnabled != nil && !tp {
-		return nil, errors.Wrap(ErrConfigValidation, "enableMetricsServer is a TechPreview feature, to be able to deploy it please enable TechPreview")
+	if c.ClusterMonitoringConfiguration.MetricsServerConfig != nil && !tp {
+		return nil, errors.Wrap(ErrConfigValidation, "Metrics Server is a TechPreview feature, to be able to deploy it please enable TechPreview")
 	}
 
 	return res, nil
@@ -325,6 +326,9 @@ func (c *Config) SetImages(images map[string]string) {
 	c.Images.OpenShiftStateMetrics = images["openshift-state-metrics"]
 	c.Images.Thanos = images["thanos"]
 	c.Images.MonitoringPlugin = images["monitoring-plugin"]
+	// Till image is updated in CVO
+	// c.Images.MetricsServer = images["metrics-server"]
+	c.Images.MetricsServer = "registry.k8s.io/metrics-server/metrics-server:v0.6.3"
 }
 
 func (c *Config) SetTelemetryMatches(matches []string) {

pkg/manifests/manifests.go

@@ -1801,6 +1801,76 @@ func (f *Factory) PrometheusUserWorkloadPrometheusServiceMonitor() (*monv1.Servi
 	return f.NewServiceMonitor(f.assets.MustNewAssetReader(PrometheusUserWorkloadPrometheusServiceMonitor))
 }
 
+func (f *Factory) MetricsServerServiceAccount() (*v1.ServiceAccount, error) {
+	return f.NewServiceAccount(f.assets.MustNewAssetReader(MetricsServerServiceAccount))
+}
+
+func (f *Factory) MetricsServerClusterRole() (*rbacv1.ClusterRole, error) {
+	return f.NewClusterRole(f.assets.MustNewAssetReader(MetricsServerClusterRole))
+}
+
+func (f *Factory) MetricsServerClusterRoleBinding() (*rbacv1.ClusterRoleBinding, error) {
+	return f.NewClusterRoleBinding(f.assets.MustNewAssetReader(MetricsServerClusterRoleBinding))
+}
+
+func (f *Factory) MetricsServerClusterRoleAggregatedMetricsReader() (*rbacv1.ClusterRole, error) {
+	return f.NewClusterRole(f.assets.MustNewAssetReader(MetricsServerClusterRoleAggregatedMetricsReader))
+}
+
+func (f *Factory) MetricsServerClusterRoleBindingAuthDelegator() (*rbacv1.ClusterRoleBinding, error) {
+	return f.NewClusterRoleBinding(f.assets.MustNewAssetReader(MetricsServerClusterRoleBindingAuthDelegator))
+}
+
+func (f *Factory) MetricsServerRoleBindingAuthReader() (*rbacv1.RoleBinding, error) {
+	return f.NewRoleBinding(f.assets.MustNewAssetReader(MetricsServerRoleBindingAuthReader))
+}
+
+func (f *Factory) MetricsServerDeployment() (*appsv1.Deployment, error) {
+	d, err := f.NewDeployment(f.assets.MustNewAssetReader(MetricsServerDeployment))
+	if err != nil {
+		return nil, err
+	}
+	for i, container := range d.Spec.Template.Spec.Containers {
+		switch container.Name {
+		case "metrics-server":
+			d.Spec.Template.Spec.Containers[i].Image = f.config.Images.MetricsServer
+
+			if f.config.ClusterMonitoringConfiguration.MetricsServerConfig != nil {
+				if f.config.ClusterMonitoringConfiguration.MetricsServerConfig.Resources != nil {
+					d.Spec.Template.Spec.Containers[i].Resources = *f.config.ClusterMonitoringConfiguration.MetricsServerConfig.Resources
+				}
+
+				if f.config.ClusterMonitoringConfiguration.MetricsServerConfig.NodeSelector != nil {
+					d.Spec.Template.Spec.NodeSelector = f.config.ClusterMonitoringConfiguration.MetricsServerConfig.NodeSelector
+				}
+
+				if f.config.ClusterMonitoringConfiguration.MetricsServerConfig.Tolerations != nil {
+					d.Spec.Template.Spec.Tolerations = f.config.ClusterMonitoringConfiguration.MetricsServerConfig.Tolerations
+				}
+			}
+
+		}
+	}
+
+	return d, nil
+}
+
+func (f *Factory) MetricsServerPodDisruptionBudget() (*policyv1.PodDisruptionBudget, error) {
+	return f.NewPodDisruptionBudget(f.assets.MustNewAssetReader(MetricsServerPodDisruptionBudget))
+}
+
+func (f *Factory) MetricsServerService() (*v1.Service, error) {
+	return f.NewService(f.assets.MustNewAssetReader(MetricsServerService))
+}
+
+func (f *Factory) MetricsServerServiceMonitor() (*monv1.ServiceMonitor, error) {
+	return f.NewServiceMonitor(f.assets.MustNewAssetReader(MetricsServerServiceMonitor))
+}
+
+func (f *Factory) MetricsServerAPIService() (*apiregistrationv1.APIService, error) {
+	return f.NewAPIService(f.assets.MustNewAssetReader(MetricsServerAPIService))
+}
+
 func (f *Factory) PrometheusAdapterClusterRole() (*rbacv1.ClusterRole, error) {
 	return f.NewClusterRole(f.assets.MustNewAssetReader(PrometheusAdapterClusterRole))
 }

pkg/manifests/types.go

@@ -43,10 +43,8 @@ type ClusterMonitoringConfiguration struct {
 	UserWorkloadEnabled *bool `json:"enableUserWorkload,omitempty"`
 	// OmitFromDoc
 	HTTPConfig *HTTPConfig `json:"http,omitempty"`
-	// `MetricsServerEnabled` is a Boolean flag that enables metrics-server.
-	MetricsServerEnabled *bool `json:"enableMetricsServer,omitempty"`
 	// `MetricsServer` defines settings for the MetricsServer component.
-	MetricsServer *MetricsServer `json:"metricsServer,omitempty"`
+	MetricsServerConfig *MetricsServerConfig `json:"metricsServer,omitempty"`
 	// `K8sPrometheusAdapter` defines settings for the Prometheus Adapter component.
 	K8sPrometheusAdapter *K8sPrometheusAdapter `json:"k8sPrometheusAdapter,omitempty"`
 	// `KubeStateMetricsConfig` defines settings for the `kube-state-metrics` agent.
@@ -105,13 +103,15 @@ type AlertmanagerMainConfig struct {
 	VolumeClaimTemplate *monv1.EmbeddedPersistentVolumeClaim `json:"volumeClaimTemplate,omitempty"`
 }
 
-// The `MetricsServer` resource defines settings for the MetricsServer component.
-type MetricsServer struct {
+// The `MetricsServerConfig` resource defines settings for the MetricsServer component.
+type MetricsServerConfig struct {
+	// `Enabled` is a Boolean flag that enables or disables metrics-server.
+	Enabled *bool `json:"enabled,omitempty"`
 	// Defines the nodes on which the pods are scheduled.
 	NodeSelector map[string]string `json:"nodeSelector,omitempty"`
 	// Defines tolerations for the pods.
 	Tolerations []v1.Toleration `json:"tolerations,omitempty"`
-	// Defines resource requests and limits for the Prometheus container.
+	// Defines resource requests and limits for the Metrics Server container.
 	Resources *v1.ResourceRequirements `json:"resources,omitempty"`
 }
 

pkg/operator/operator.go

@@ -742,6 +742,7 @@ func (o *Operator) sync(ctx context.Context, key string) error {
 				tasks.NewTaskSpec("Updating kube-state-metrics", tasks.NewKubeStateMetricsTask(o.client, factory)),
 				tasks.NewTaskSpec("Updating openshift-state-metrics", tasks.NewOpenShiftStateMetricsTask(o.client, factory)),
 				tasks.NewTaskSpec("Updating prometheus-adapter", tasks.NewPrometheusAdapterTask(ctx, o.namespace, o.client, factory, config)),
+				tasks.NewTaskSpec("Updating metrics-server", tasks.NewMetricsServerTask(ctx, o.namespace, o.client, factory, config)),
 				tasks.NewTaskSpec("Updating Telemeter client", tasks.NewTelemeterClientTask(o.client, factory, config)),
 				tasks.NewTaskSpec("Updating Thanos Querier", tasks.NewThanosQuerierTask(o.client, factory, config)),
 				tasks.NewTaskSpec("Updating User Workload Thanos Ruler", tasks.NewThanosRulerUserWorkloadTask(o.client, factory, config)),