Skip to main content
Version: 1.3.0

Slice Operator Metrics

The Slice Operator metrics provide valuable insights into the performance and health of various components within the KubeSlice environment. These metrics are categorized based on different labels associated with each component. The metrics include information such as the number of active endpoints in service imports and exports, the number of application pods in each slice per namespace, the health status of the worker cluster and its components, the health status of individual slices and their components, and counters for slice-related events. These metrics help you monitor the state of your KubeSlice deployment and make informed decisions for efficient management. The metrics are accompanied by labels that provide additional context, such as the action triggering the metric, the component or object involved, the project and cluster names, and more. By leveraging these metrics and labels, you can effectively monitor and optimize your KubeSlice infrastructure.

Sl. NoNameTypeLabelsDescription
1kubeslice_serviceimport_endpointsGaugeslice_project, slice_cluster, slice, namespace, service, slice_reporting_controllerNumber of active endpoints in the service import.
2kubeslice_serviceexport_endpointsGaugeslice_project, slice_cluster, slice, namespace, service, slice_reporting_controllerNumber of active endpoints in the service export.
3kubeslice_app_podsGaugeslice_project, slice_cluster, slice, slice_namespaceNumber of application pods in each slice per namespace.
4kubeslice_cluster_upGaugeslice_project, slice_clusterDenotes 1 if the worker cluster is healthy and denotes 0 if it is unhealthy.
5kubeslice_cluster_component_upGaugeslice_project, slice_cluster, componentDenotes 1 if the worker cluster is healthy and denotes 0 if it is unhealthy.
6kubeslice_slice_upGaugeslice_project, slice_cluster, sliceDenotes 1 if the slice is healthy and denotes 0 if it is unhealthy.
7kubeslice_slice_component_upGaugeslice_project, slice_cluster, slice, componentDenotes 1 if the slice component is healthy and denotes 0 if it is unhealthy.
8kubeslice_slice_created_total, kubeslice_slice_updated_total, kubeslice_slice_deleted_total, kubeslice_slice_creation_failed_total, kubeslice_slice_updation_failed_total, kubeslice_slice_deletion_failed_totalCounterslice_project, slice_cluster, sliceCounter for slice related events.
9kubeslice_nodeaffinity_taints_activeGaugeslice_project, slice_cluster, slice, namespace, slice_nodeaffinityNumber of nodes tainted for Node Affinity in the slice.
10kubeslice_nodeaffinity_mutated_workloads_totalCounterslice_project, slice_cluster, slice, namespace, slice_nodeaffinity, slice_workload_typeCount of workloads mutated by webhook for Node Affinity.

Labels

The following table describes the labels that are attached to the Slice Operator metrics.

LabelDescription
actionThe action which triggered the metric record.
componentThe KubeSlice component to which the metric is attached to.
eventThe event title to which the metric is attached to.
object_kindThe kind of the object on which the metric is recorded.
object_nameThe name of the object on which the metric is recorded.
namespaceThe namespace of the object for which the metric is recorded.
serviceThe metric is related to the service discovery (service export or import).
sliceThe name of the slice which the metric is attached to. The value is NA if the metric is unrelated to any slice.
slice_clusterThe name of the cluster the metric is attached to. The value is worker-cluster for all the Worker Operator related metrics.
slice_gatewayThe Slice Gateway for which the metric is recorded
slice_gateway_podThe Slice Gateway pod for which the metric is recorded.
slice_nodeaffinityThe Node Affinity of the slice for which the metric is recorded.
slice_networkpolicyThe namespace isolation policy of the slice for which the metric is recorded.
slice_networkpolicy_violationThe namespace isolation policy violation of the slice for which the metric is recorded.
slice_projectThe name of the project on which the slice is created.
slice_reporting_controllerThe name of the reporting controller. The value is controller for all the KubeSlice Controller related metrics.
slice_workload_typeThe workload type of the slice for which the metric is recorded.