Documentation
¶
Index ¶
- func DevicePluginDaemonset() *apps.DaemonSet
- func DynamicResourceAllocationClusterRole() *rbac.ClusterRole
- func DynamicResourceAllocationClusterRoleBinding() *rbac.ClusterRoleBinding
- func DynamicResourceAllocationDaemonset() *apps.DaemonSet
- func DynamicResourceAllocationDeviceClass(limitToKMDs bool) *resv1.DeviceClass
- func DynamicResourceAllocationDeviceClassVfio(limitToVfio bool) *resv1.DeviceClass
- func DynamicResourceAllocationMonitorClaimTemplate() *resv1.ResourceClaimTemplate
- func DynamicResourceAllocationServiceAccount() *core.ServiceAccount
- func DynamicResourceAllocationValidatingAdmissionPolicy() *adreg.ValidatingAdmissionPolicy
- func DynamicResourceAllocationValidatingAdmissionPolicyBinding() *adreg.ValidatingAdmissionPolicyBinding
- func NFDNodeFeatureRulesGpu() *nfdcrd.NodeFeatureRule
- func PrometheusServiceMonitor() *prometheusv1.ServiceMonitor
- func XpuManagerDaemonset() *apps.DaemonSet
- func XpuManagerFWUpdateJob() *batch.Job
- func XpuManagerResetJob() *batch.Job
- func XpuManagerService() *core.Service
- type HWSeverityMapping
- type HWStatusMapping
- type IntelXPUInfoExporter
- type IntelXPUReceiver
- type IntelXPUStatusProcessor
- type KeyValues
- type OTelConfig
- type OTelExporters
- type OTelLogs
- type OTelPipelines
- type OTelProcessors
- type OTelReceivers
- type OTelService
- type OTelTelemetry
- type PipelineLogs
- type PipelineMetrics
- type PrometheusExporter
- type StateCondition
- type StatusRule
- type StatusState
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func DevicePluginDaemonset ¶
func DynamicResourceAllocationClusterRole ¶
func DynamicResourceAllocationClusterRole() *rbac.ClusterRole
func DynamicResourceAllocationClusterRoleBinding ¶
func DynamicResourceAllocationClusterRoleBinding() *rbac.ClusterRoleBinding
func DynamicResourceAllocationDeviceClass ¶
func DynamicResourceAllocationDeviceClass(limitToKMDs bool) *resv1.DeviceClass
func DynamicResourceAllocationDeviceClassVfio ¶ added in v0.3.0
func DynamicResourceAllocationDeviceClassVfio(limitToVfio bool) *resv1.DeviceClass
func DynamicResourceAllocationMonitorClaimTemplate ¶
func DynamicResourceAllocationMonitorClaimTemplate() *resv1.ResourceClaimTemplate
func DynamicResourceAllocationServiceAccount ¶
func DynamicResourceAllocationServiceAccount() *core.ServiceAccount
func DynamicResourceAllocationValidatingAdmissionPolicy ¶
func DynamicResourceAllocationValidatingAdmissionPolicy() *adreg.ValidatingAdmissionPolicy
func DynamicResourceAllocationValidatingAdmissionPolicyBinding ¶
func DynamicResourceAllocationValidatingAdmissionPolicyBinding() *adreg.ValidatingAdmissionPolicyBinding
func NFDNodeFeatureRulesGpu ¶
func NFDNodeFeatureRulesGpu() *nfdcrd.NodeFeatureRule
func PrometheusServiceMonitor ¶
func PrometheusServiceMonitor() *prometheusv1.ServiceMonitor
func XpuManagerDaemonset ¶
func XpuManagerFWUpdateJob ¶
func XpuManagerResetJob ¶ added in v0.4.0
func XpuManagerService ¶
Types ¶
type HWSeverityMapping ¶ added in v0.3.0
type HWSeverityMapping struct {
Severity string `json:"severity"`
Message string `json:"message,omitempty"`
}
HWSeverityMapping maps a health state name to a severity level and optional message.
type HWStatusMapping ¶ added in v0.3.0
type HWStatusMapping struct {
HealthDomain string `json:"health_domain"`
Filters []KeyValues `json:"filters,omitempty"`
StateMapping map[string]HWSeverityMapping `json:"state_mapping"`
}
HWStatusMapping maps a health domain to state severity entries.
The embedded otel-config.yaml is parsed into these types and re-marshalled into the ConfigMap, so a key missing here is silently dropped from what xpumd actually reads.
type IntelXPUInfoExporter ¶
type IntelXPUInfoExporter struct {
Endpoint string `json:"endpoint,omitempty"`
HWStatusMappings []HWStatusMapping `json:"hw_status_mappings,omitempty"`
}
IntelXPUInfoExporter exports health status over a unix domain socket.
type IntelXPUReceiver ¶
type IntelXPUReceiver struct {
CollectionInterval string `json:"collection_interval"`
InitialDelay string `json:"initial_delay"`
Timeout int `json:"timeout"`
SamplingInterval string `json:"sampling_interval"`
Metrics map[string]interface{} `json:"metrics,omitempty"`
}
IntelXPUReceiver is the configuration for the intel_xpu receiver.
type IntelXPUStatusProcessor ¶
type IntelXPUStatusProcessor struct {
Rules []StatusRule `json:"rules"`
}
IntelXPUStatusProcessor is the configuration for the intel_xpu_status processor.
type OTelConfig ¶
type OTelConfig struct {
Receivers OTelReceivers `json:"receivers"`
Processors OTelProcessors `json:"processors"`
Exporters OTelExporters `json:"exporters"`
Extensions map[string]interface{} `json:"extensions,omitempty"`
Service OTelService `json:"service"`
}
OTelConfig represents the OpenTelemetry Collector configuration for XPU Manager.
func XpuManagerOTelConfig ¶
func XpuManagerOTelConfig() *OTelConfig
type OTelExporters ¶
type OTelExporters struct {
IntelXPUInfo IntelXPUInfoExporter `json:"intel_xpu_info"`
Prometheus PrometheusExporter `json:"prometheus"`
}
OTelExporters holds exporter configurations.
type OTelLogs ¶
type OTelLogs struct {
Level string `json:"level"`
}
OTelLogs configures the log level for the collector itself.
type OTelPipelines ¶
type OTelPipelines struct {
Metrics PipelineMetrics `json:"metrics"`
Logs *PipelineLogs `json:"logs,omitempty"`
}
OTelPipelines defines the data pipelines in the collector.
type OTelProcessors ¶
type OTelProcessors struct {
IntelXPUStatus IntelXPUStatusProcessor `json:"intel_xpu_status"`
}
OTelProcessors holds processor configurations.
type OTelReceivers ¶
type OTelReceivers struct {
IntelXPU IntelXPUReceiver `json:"intel_xpu"`
}
OTelReceivers holds receiver configurations.
type OTelService ¶
type OTelService struct {
Telemetry OTelTelemetry `json:"telemetry"`
Pipelines OTelPipelines `json:"pipelines"`
}
OTelService is the top-level service configuration block.
type OTelTelemetry ¶
type OTelTelemetry struct {
Logs OTelLogs `json:"logs"`
}
OTelTelemetry configures the collector's own telemetry output.
type PipelineLogs ¶ added in v0.3.0
type PipelineLogs struct {
Receivers []string `json:"receivers"`
Processors []string `json:"processors,omitempty"`
Exporters []string `json:"exporters"`
}
PipelineLogs lists the components forming the logs pipeline.
type PipelineMetrics ¶
type PipelineMetrics struct {
Receivers []string `json:"receivers"`
Processors []string `json:"processors"`
Exporters []string `json:"exporters"`
}
PipelineMetrics lists the components forming the metrics pipeline.
type PrometheusExporter ¶
type PrometheusExporter struct {
Endpoint string `json:"endpoint"`
}
PrometheusExporter exposes metrics on a Prometheus-compatible HTTP endpoint.
type StateCondition ¶
type StateCondition struct {
Value float64 `json:"value"`
ParentFilters []KeyValues `json:"parent_filters,omitempty"`
}
StateCondition is a threshold condition for a health state, with optional device filters.
type StatusRule ¶
type StatusRule struct {
Name string `json:"name"`
SourceMetric string `json:"source_metric"`
ParentMetric string `json:"parent_metric"`
ParentRefAttribute string `json:"parent_ref_attribute"`
ParentFilters []KeyValues `json:"parent_filters"`
ComponentFilters []KeyValues `json:"component_filters"`
CopyAttributes []string `json:"copy_attributes"`
AddAttributes map[string]string `json:"add_attributes"`
States []StatusState `json:"states"`
}
StatusRule defines a single health evaluation rule.
type StatusState ¶
type StatusState struct {
StateName string `json:"state_name"`
Conditions []StateCondition `json:"conditions,omitempty"`
}
StatusState defines a named health state, optionally gated by conditions.