Sign in Dashboard Get started
  1. Registry
  2. Packages
  3. CAST AI
  4. API Docs
  5. WorkloadScalingPolicy
Viewing docs for CAST AI v8.60.0
published on Wednesday, Aug 19, 2026 by CAST AI
Schema (JSON)
castai/pulumi-castai

castai.WorkloadScalingPolicy

castai logo
Viewing docs for CAST AI v8.60.0
published on Wednesday, Aug 19, 2026 by CAST AI
Schema (JSON)
castai/pulumi-castai

    Create WorkloadScalingPolicy Resource

    Resources are created with functions called constructors. To learn more about declaring and configuring resources, see Resources.

    Constructor syntax

    new WorkloadScalingPolicy(name: string, args: WorkloadScalingPolicyArgs, opts?: CustomResourceOptions);
    @overload
def WorkloadScalingPolicy(resource_name: str,
                          args: WorkloadScalingPolicyArgs,
                          opts: Optional[ResourceOptions] = None)

@overload
def WorkloadScalingPolicy(resource_name: str,
                          opts: Optional[ResourceOptions] = None,
                          memory: Optional[_workload.WorkloadScalingPolicyMemoryArgs] = None,
                          cpu: Optional[_workload.WorkloadScalingPolicyCpuArgs] = None,
                          apply_type: Optional[str] = None,
                          management_option: Optional[str] = None,
                          cluster_id: Optional[str] = None,
                          jvm: Optional[_workload.WorkloadScalingPolicyJvmArgs] = None,
                          confidence: Optional[_workload.WorkloadScalingPolicyConfidenceArgs] = None,
                          downscaling: Optional[_workload.WorkloadScalingPolicyDownscalingArgs] = None,
                          anti_affinity: Optional[_workload.WorkloadScalingPolicyAntiAffinityArgs] = None,
                          excluded_containers: Optional[Sequence[str]] = None,
                          hpa_converters: Optional[Sequence[_workload.WorkloadScalingPolicyHpaConverterArgs]] = None,
                          assignment_rules: Optional[Sequence[_workload.WorkloadScalingPolicyAssignmentRuleArgs]] = None,
                          anomaly_detection: Optional[_workload.WorkloadScalingPolicyAnomalyDetectionArgs] = None,
                          memory_event: Optional[_workload.WorkloadScalingPolicyMemoryEventArgs] = None,
                          name: Optional[str] = None,
                          predictive_scaling: Optional[_workload.WorkloadScalingPolicyPredictiveScalingArgs] = None,
                          rollout_behavior: Optional[_workload.WorkloadScalingPolicyRolloutBehaviorArgs] = None,
                          startup: Optional[_workload.WorkloadScalingPolicyStartupArgs] = None)
    func NewWorkloadScalingPolicy(ctx *Context, name string, args WorkloadScalingPolicyArgs, opts ...ResourceOption) (*WorkloadScalingPolicy, error)
    public WorkloadScalingPolicy(string name, WorkloadScalingPolicyArgs args, CustomResourceOptions? opts = null)
    public WorkloadScalingPolicy(String name, WorkloadScalingPolicyArgs args)
public WorkloadScalingPolicy(String name, WorkloadScalingPolicyArgs args, CustomResourceOptions options)

    type: castai:workload:WorkloadScalingPolicy
properties: # The arguments to resource properties.
options: # Bag of options to control resource's behavior.


    resource "castai_workload_scaling_policy" "name" {
    # resource properties
}

    Parameters

    name string
    The unique name of the resource.
    args WorkloadScalingPolicyArgs
    The arguments to resource properties.
    opts CustomResourceOptions
    Bag of options to control resource's behavior.
    resource_name str
    The unique name of the resource.
    args WorkloadScalingPolicyArgs
    The arguments to resource properties.
    opts ResourceOptions
    Bag of options to control resource's behavior.
    ctx Context
    Context object for the current deployment.
    name string
    The unique name of the resource.
    args WorkloadScalingPolicyArgs
    The arguments to resource properties.
    opts ResourceOption
    Bag of options to control resource's behavior.
    name string
    The unique name of the resource.
    args WorkloadScalingPolicyArgs
    The arguments to resource properties.
    opts CustomResourceOptions
    Bag of options to control resource's behavior.
    name String
    The unique name of the resource.
    args WorkloadScalingPolicyArgs
    The arguments to resource properties.
    options CustomResourceOptions
    Bag of options to control resource's behavior.

    WorkloadScalingPolicy Resource Properties

    To learn more about resource properties and how to use them, see Inputs and Outputs in the Architecture and Concepts docs.

    Inputs

    In Python, inputs that are objects can be passed either as argument classes or as dictionary literals.

    The WorkloadScalingPolicy resource accepts the following input properties:

    ApplyType string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    ClusterId string
    CAST AI cluster id
    Cpu Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyCpu
    ManagementOption string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    Memory Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemory
    AnomalyDetection Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAnomalyDetection
    Defines anomaly detection settings for the scaling policy.
    AntiAffinity Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAntiAffinity
    AssignmentRules List<Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAssignmentRule>
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    Confidence Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyConfidence
    Defines the confidence settings for applying recommendations.
    Downscaling Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyDownscaling
    ExcludedContainers List<string>
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    HpaConverters List<Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyHpaConverter>
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    Jvm Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyJvm
    JVM optimization settings.
    MemoryEvent Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemoryEvent
    Name string
    Scaling policy name
    PredictiveScaling Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyPredictiveScaling
    RolloutBehavior Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyRolloutBehavior
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    Startup Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyStartup
    ApplyType string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    ClusterId string
    CAST AI cluster id
    Cpu WorkloadScalingPolicyCpuArgs
    ManagementOption string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    Memory WorkloadScalingPolicyMemoryArgs
    AnomalyDetection WorkloadScalingPolicyAnomalyDetectionArgs
    Defines anomaly detection settings for the scaling policy.
    AntiAffinity WorkloadScalingPolicyAntiAffinityArgs
    AssignmentRules WorkloadScalingPolicyAssignmentRuleArgs
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    Confidence WorkloadScalingPolicyConfidenceArgs
    Defines the confidence settings for applying recommendations.
    Downscaling WorkloadScalingPolicyDownscalingArgs
    ExcludedContainers []string
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    HpaConverters WorkloadScalingPolicyHpaConverterArgs
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    Jvm WorkloadScalingPolicyJvmArgs
    JVM optimization settings.
    MemoryEvent WorkloadScalingPolicyMemoryEventArgs
    Name string
    Scaling policy name
    PredictiveScaling WorkloadScalingPolicyPredictiveScalingArgs
    RolloutBehavior WorkloadScalingPolicyRolloutBehaviorArgs
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    Startup WorkloadScalingPolicyStartupArgs
    apply_type string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    cluster_id string
    CAST AI cluster id
    cpu object
    management_option string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory object
    anomaly_detection object
    Defines anomaly detection settings for the scaling policy.
    anti_affinity object
    assignment_rules list(object)
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    confidence object
    Defines the confidence settings for applying recommendations.
    downscaling object
    excluded_containers list(string)
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpa_converters list(object)
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm object
    JVM optimization settings.
    memory_event object
    name string
    Scaling policy name
    predictive_scaling object
    rollout_behavior object
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup object
    applyType String
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    clusterId String
    CAST AI cluster id
    cpu WorkloadScalingPolicyCpu
    managementOption String
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory WorkloadScalingPolicyMemory
    anomalyDetection WorkloadScalingPolicyAnomalyDetection
    Defines anomaly detection settings for the scaling policy.
    antiAffinity WorkloadScalingPolicyAntiAffinity
    assignmentRules List<WorkloadScalingPolicyAssignmentRule>
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    confidence WorkloadScalingPolicyConfidence
    Defines the confidence settings for applying recommendations.
    downscaling WorkloadScalingPolicyDownscaling
    excludedContainers List<String>
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpaConverters List<WorkloadScalingPolicyHpaConverter>
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm WorkloadScalingPolicyJvm
    JVM optimization settings.
    memoryEvent WorkloadScalingPolicyMemoryEvent
    name String
    Scaling policy name
    predictiveScaling WorkloadScalingPolicyPredictiveScaling
    rolloutBehavior WorkloadScalingPolicyRolloutBehavior
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup WorkloadScalingPolicyStartup
    applyType string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    clusterId string
    CAST AI cluster id
    cpu workloadWorkloadScalingPolicyCpu
    managementOption string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory workloadWorkloadScalingPolicyMemory
    anomalyDetection workloadWorkloadScalingPolicyAnomalyDetection
    Defines anomaly detection settings for the scaling policy.
    antiAffinity workloadWorkloadScalingPolicyAntiAffinity
    assignmentRules workloadWorkloadScalingPolicyAssignmentRule[]
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    confidence workloadWorkloadScalingPolicyConfidence
    Defines the confidence settings for applying recommendations.
    downscaling workloadWorkloadScalingPolicyDownscaling
    excludedContainers string[]
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpaConverters workloadWorkloadScalingPolicyHpaConverter[]
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm workloadWorkloadScalingPolicyJvm
    JVM optimization settings.
    memoryEvent workloadWorkloadScalingPolicyMemoryEvent
    name string
    Scaling policy name
    predictiveScaling workloadWorkloadScalingPolicyPredictiveScaling
    rolloutBehavior workloadWorkloadScalingPolicyRolloutBehavior
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup workloadWorkloadScalingPolicyStartup
    apply_type str
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    cluster_id str
    CAST AI cluster id
    cpu workload.WorkloadScalingPolicyCpuArgs
    management_option str
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory workload.WorkloadScalingPolicyMemoryArgs
    anomaly_detection workload.WorkloadScalingPolicyAnomalyDetectionArgs
    Defines anomaly detection settings for the scaling policy.
    anti_affinity workload.WorkloadScalingPolicyAntiAffinityArgs
    assignment_rules Sequence[workload.WorkloadScalingPolicyAssignmentRuleArgs]
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    confidence workload.WorkloadScalingPolicyConfidenceArgs
    Defines the confidence settings for applying recommendations.
    downscaling workload.WorkloadScalingPolicyDownscalingArgs
    excluded_containers Sequence[str]
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpa_converters Sequence[workload.WorkloadScalingPolicyHpaConverterArgs]
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm workload.WorkloadScalingPolicyJvmArgs
    JVM optimization settings.
    memory_event workload.WorkloadScalingPolicyMemoryEventArgs
    name str
    Scaling policy name
    predictive_scaling workload.WorkloadScalingPolicyPredictiveScalingArgs
    rollout_behavior workload.WorkloadScalingPolicyRolloutBehaviorArgs
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup workload.WorkloadScalingPolicyStartupArgs
    applyType String
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    clusterId String
    CAST AI cluster id
    cpu Property Map
    managementOption String
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory Property Map
    anomalyDetection Property Map
    Defines anomaly detection settings for the scaling policy.
    antiAffinity Property Map
    assignmentRules List<Property Map>
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    confidence Property Map
    Defines the confidence settings for applying recommendations.
    downscaling Property Map
    excludedContainers List<String>
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpaConverters List<Property Map>
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm Property Map
    JVM optimization settings.
    memoryEvent Property Map
    name String
    Scaling policy name
    predictiveScaling Property Map
    rolloutBehavior Property Map
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup Property Map

    Outputs

    All input properties are implicitly available as output properties. Additionally, the WorkloadScalingPolicy resource produces the following output properties:

    Id string
    The provider-assigned unique ID for this managed resource.
    Id string
    The provider-assigned unique ID for this managed resource.
    id string
    The provider-assigned unique ID for this managed resource.
    id String
    The provider-assigned unique ID for this managed resource.
    id string
    The provider-assigned unique ID for this managed resource.
    id str
    The provider-assigned unique ID for this managed resource.
    id String
    The provider-assigned unique ID for this managed resource.

    Look up Existing WorkloadScalingPolicy Resource

    Get an existing WorkloadScalingPolicy resource’s state with the given name, ID, and optional extra properties used to qualify the lookup.

    public static get(name: string, id: Input<ID>, state?: WorkloadScalingPolicyState, opts?: CustomResourceOptions): WorkloadScalingPolicy
    @staticmethod
def get(resource_name: str,
        id: str,
        opts: Optional[ResourceOptions] = None,
        anomaly_detection: Optional[_workload.WorkloadScalingPolicyAnomalyDetectionArgs] = None,
        anti_affinity: Optional[_workload.WorkloadScalingPolicyAntiAffinityArgs] = None,
        apply_type: Optional[str] = None,
        assignment_rules: Optional[Sequence[_workload.WorkloadScalingPolicyAssignmentRuleArgs]] = None,
        cluster_id: Optional[str] = None,
        confidence: Optional[_workload.WorkloadScalingPolicyConfidenceArgs] = None,
        cpu: Optional[_workload.WorkloadScalingPolicyCpuArgs] = None,
        downscaling: Optional[_workload.WorkloadScalingPolicyDownscalingArgs] = None,
        excluded_containers: Optional[Sequence[str]] = None,
        hpa_converters: Optional[Sequence[_workload.WorkloadScalingPolicyHpaConverterArgs]] = None,
        jvm: Optional[_workload.WorkloadScalingPolicyJvmArgs] = None,
        management_option: Optional[str] = None,
        memory: Optional[_workload.WorkloadScalingPolicyMemoryArgs] = None,
        memory_event: Optional[_workload.WorkloadScalingPolicyMemoryEventArgs] = None,
        name: Optional[str] = None,
        predictive_scaling: Optional[_workload.WorkloadScalingPolicyPredictiveScalingArgs] = None,
        rollout_behavior: Optional[_workload.WorkloadScalingPolicyRolloutBehaviorArgs] = None,
        startup: Optional[_workload.WorkloadScalingPolicyStartupArgs] = None) -> WorkloadScalingPolicy
    func GetWorkloadScalingPolicy(ctx *Context, name string, id IDInput, state *WorkloadScalingPolicyState, opts ...ResourceOption) (*WorkloadScalingPolicy, error)
    public static WorkloadScalingPolicy Get(string name, Input<string> id, WorkloadScalingPolicyState? state, CustomResourceOptions? opts = null)
    public static WorkloadScalingPolicy get(String name, Output<String> id, WorkloadScalingPolicyState state, CustomResourceOptions options)
    resources:  _:    type: castai:workload:WorkloadScalingPolicy    get:      id: ${id}
    import {
  to = castai_workload_scaling_policy.example
  id = "${id}"
}
    name
    The unique name of the resulting resource.
    id
    The unique provider ID of the resource to lookup.
    state
    Any extra arguments used during the lookup.
    opts
    A bag of options that control this resource's behavior.
    resource_name
    The unique name of the resulting resource.
    id
    The unique provider ID of the resource to lookup.
    name
    The unique name of the resulting resource.
    id
    The unique provider ID of the resource to lookup.
    state
    Any extra arguments used during the lookup.
    opts
    A bag of options that control this resource's behavior.
    name
    The unique name of the resulting resource.
    id
    The unique provider ID of the resource to lookup.
    state
    Any extra arguments used during the lookup.
    opts
    A bag of options that control this resource's behavior.
    name
    The unique name of the resulting resource.
    id
    The unique provider ID of the resource to lookup.
    state
    Any extra arguments used during the lookup.
    opts
    A bag of options that control this resource's behavior.
    The following state arguments are supported:
    AnomalyDetection Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAnomalyDetection
    Defines anomaly detection settings for the scaling policy.
    AntiAffinity Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAntiAffinity
    ApplyType string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    AssignmentRules List<Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAssignmentRule>
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    ClusterId string
    CAST AI cluster id
    Confidence Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyConfidence
    Defines the confidence settings for applying recommendations.
    Cpu Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyCpu
    Downscaling Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyDownscaling
    ExcludedContainers List<string>
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    HpaConverters List<Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyHpaConverter>
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    Jvm Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyJvm
    JVM optimization settings.
    ManagementOption string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    Memory Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemory
    MemoryEvent Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemoryEvent
    Name string
    Scaling policy name
    PredictiveScaling Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyPredictiveScaling
    RolloutBehavior Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyRolloutBehavior
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    Startup Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyStartup
    AnomalyDetection WorkloadScalingPolicyAnomalyDetectionArgs
    Defines anomaly detection settings for the scaling policy.
    AntiAffinity WorkloadScalingPolicyAntiAffinityArgs
    ApplyType string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    AssignmentRules WorkloadScalingPolicyAssignmentRuleArgs
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    ClusterId string
    CAST AI cluster id
    Confidence WorkloadScalingPolicyConfidenceArgs
    Defines the confidence settings for applying recommendations.
    Cpu WorkloadScalingPolicyCpuArgs
    Downscaling WorkloadScalingPolicyDownscalingArgs
    ExcludedContainers []string
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    HpaConverters WorkloadScalingPolicyHpaConverterArgs
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    Jvm WorkloadScalingPolicyJvmArgs
    JVM optimization settings.
    ManagementOption string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    Memory WorkloadScalingPolicyMemoryArgs
    MemoryEvent WorkloadScalingPolicyMemoryEventArgs
    Name string
    Scaling policy name
    PredictiveScaling WorkloadScalingPolicyPredictiveScalingArgs
    RolloutBehavior WorkloadScalingPolicyRolloutBehaviorArgs
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    Startup WorkloadScalingPolicyStartupArgs
    anomaly_detection object
    Defines anomaly detection settings for the scaling policy.
    anti_affinity object
    apply_type string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    assignment_rules list(object)
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    cluster_id string
    CAST AI cluster id
    confidence object
    Defines the confidence settings for applying recommendations.
    cpu object
    downscaling object
    excluded_containers list(string)
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpa_converters list(object)
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm object
    JVM optimization settings.
    management_option string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory object
    memory_event object
    name string
    Scaling policy name
    predictive_scaling object
    rollout_behavior object
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup object
    anomalyDetection WorkloadScalingPolicyAnomalyDetection
    Defines anomaly detection settings for the scaling policy.
    antiAffinity WorkloadScalingPolicyAntiAffinity
    applyType String
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    assignmentRules List<WorkloadScalingPolicyAssignmentRule>
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    clusterId String
    CAST AI cluster id
    confidence WorkloadScalingPolicyConfidence
    Defines the confidence settings for applying recommendations.
    cpu WorkloadScalingPolicyCpu
    downscaling WorkloadScalingPolicyDownscaling
    excludedContainers List<String>
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpaConverters List<WorkloadScalingPolicyHpaConverter>
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm WorkloadScalingPolicyJvm
    JVM optimization settings.
    managementOption String
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory WorkloadScalingPolicyMemory
    memoryEvent WorkloadScalingPolicyMemoryEvent
    name String
    Scaling policy name
    predictiveScaling WorkloadScalingPolicyPredictiveScaling
    rolloutBehavior WorkloadScalingPolicyRolloutBehavior
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup WorkloadScalingPolicyStartup
    anomalyDetection workloadWorkloadScalingPolicyAnomalyDetection
    Defines anomaly detection settings for the scaling policy.
    antiAffinity workloadWorkloadScalingPolicyAntiAffinity
    applyType string
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    assignmentRules workloadWorkloadScalingPolicyAssignmentRule[]
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    clusterId string
    CAST AI cluster id
    confidence workloadWorkloadScalingPolicyConfidence
    Defines the confidence settings for applying recommendations.
    cpu workloadWorkloadScalingPolicyCpu
    downscaling workloadWorkloadScalingPolicyDownscaling
    excludedContainers string[]
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpaConverters workloadWorkloadScalingPolicyHpaConverter[]
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm workloadWorkloadScalingPolicyJvm
    JVM optimization settings.
    managementOption string
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory workloadWorkloadScalingPolicyMemory
    memoryEvent workloadWorkloadScalingPolicyMemoryEvent
    name string
    Scaling policy name
    predictiveScaling workloadWorkloadScalingPolicyPredictiveScaling
    rolloutBehavior workloadWorkloadScalingPolicyRolloutBehavior
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup workloadWorkloadScalingPolicyStartup
    anomaly_detection workload.WorkloadScalingPolicyAnomalyDetectionArgs
    Defines anomaly detection settings for the scaling policy.
    anti_affinity workload.WorkloadScalingPolicyAntiAffinityArgs
    apply_type str
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    assignment_rules Sequence[workload.WorkloadScalingPolicyAssignmentRuleArgs]
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    cluster_id str
    CAST AI cluster id
    confidence workload.WorkloadScalingPolicyConfidenceArgs
    Defines the confidence settings for applying recommendations.
    cpu workload.WorkloadScalingPolicyCpuArgs
    downscaling workload.WorkloadScalingPolicyDownscalingArgs
    excluded_containers Sequence[str]
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpa_converters Sequence[workload.WorkloadScalingPolicyHpaConverterArgs]
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm workload.WorkloadScalingPolicyJvmArgs
    JVM optimization settings.
    management_option str
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory workload.WorkloadScalingPolicyMemoryArgs
    memory_event workload.WorkloadScalingPolicyMemoryEventArgs
    name str
    Scaling policy name
    predictive_scaling workload.WorkloadScalingPolicyPredictiveScalingArgs
    rollout_behavior workload.WorkloadScalingPolicyRolloutBehaviorArgs
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup workload.WorkloadScalingPolicyStartupArgs
    anomalyDetection Property Map
    Defines anomaly detection settings for the scaling policy.
    antiAffinity Property Map
    applyType String
    Recommendation apply type. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    assignmentRules List<Property Map>
    Allows defining conditions for automatically assigning workloads to this scaling policy.
    clusterId String
    CAST AI cluster id
    confidence Property Map
    Defines the confidence settings for applying recommendations.
    cpu Property Map
    downscaling Property Map
    excludedContainers List<String>
    Defines containers to be excluded from receiving recommendations. The containers are matched by exact name.
    hpaConverters List<Property Map>
    Configuration for converting existing HPAs when VPA is the sole optimization. If HPA management is enabled, it takes precedence over this setting.
    jvm Property Map
    JVM optimization settings.
    managementOption String
    Defines possible options for workload management. - READ_ONLY - workload watched (metrics collected), but no actions performed by CAST AI. - MANAGED - workload watched (metrics collected), CAST AI may perform actions on the workload.
    memory Property Map
    memoryEvent Property Map
    name String
    Scaling policy name
    predictiveScaling Property Map
    rolloutBehavior Property Map
    Defines the rollout behavior used when applying recommendations. Prerequisites: - Applicable to Deployment resources that support running as multi-replica. - Deployment is running with single replica (replica count = 1). - Deployment's rollout strategy allows for downtime. - Recommendation apply type is "immediate". - Cluster has workload-autoscaler component version v0.35.3 or higher.
    startup Property Map

    Supporting Types

    WorkloadScalingPolicyAnomalyDetection, WorkloadScalingPolicyAnomalyDetectionArgs

    CpuPressure WorkloadScalingPolicyAnomalyDetectionCpuPressure
    Configures CPU pressure anomaly detection thresholds.
    cpu_pressure object
    Configures CPU pressure anomaly detection thresholds.
    cpuPressure WorkloadScalingPolicyAnomalyDetectionCpuPressure
    Configures CPU pressure anomaly detection thresholds.
    cpuPressure workloadWorkloadScalingPolicyAnomalyDetectionCpuPressure
    Configures CPU pressure anomaly detection thresholds.
    cpu_pressure workload.WorkloadScalingPolicyAnomalyDetectionCpuPressure
    Configures CPU pressure anomaly detection thresholds.
    cpuPressure Property Map
    Configures CPU pressure anomaly detection thresholds.

    WorkloadScalingPolicyAnomalyDetectionCpuPressure, WorkloadScalingPolicyAnomalyDetectionCpuPressureArgs

    CpuStallThresholdPercentage double
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    MinPressuredPodPercentage double
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.
    CpuStallThresholdPercentage float64
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    MinPressuredPodPercentage float64
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.
    cpu_stall_threshold_percentage number
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    min_pressured_pod_percentage number
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.
    cpuStallThresholdPercentage Double
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    minPressuredPodPercentage Double
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.
    cpuStallThresholdPercentage number
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    minPressuredPodPercentage number
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.
    cpu_stall_threshold_percentage float
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    min_pressured_pod_percentage float
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.
    cpuStallThresholdPercentage Number
    Percentage of time (0-100) that a pod must experience CPU pressure to be considered under pressure.
    minPressuredPodPercentage Number
    Percentage (0-100) of pods that must be experiencing pressure for the detector to trigger.

    WorkloadScalingPolicyAntiAffinity, WorkloadScalingPolicyAntiAffinityArgs

    ConsiderAntiAffinity bool
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.
    ConsiderAntiAffinity bool
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.
    consider_anti_affinity bool
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.
    considerAntiAffinity Boolean
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.
    considerAntiAffinity boolean
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.
    consider_anti_affinity bool
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.
    considerAntiAffinity Boolean
    Defines if anti-affinity should be considered when scaling the workload. If enabled, requiring host ports, or having anti-affinity on hostname will force all recommendations to be deferred.

    WorkloadScalingPolicyAssignmentRule, WorkloadScalingPolicyAssignmentRuleArgs

    WorkloadScalingPolicyAssignmentRuleRule, WorkloadScalingPolicyAssignmentRuleRuleArgs

    Namespace Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAssignmentRuleRuleNamespace
    Allows assigning a scaling policy based on the workload's namespace.
    Workload Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAssignmentRuleRuleWorkload
    Allows assigning a scaling policy based on the workload's metadata.
    Namespace WorkloadScalingPolicyAssignmentRuleRuleNamespace
    Allows assigning a scaling policy based on the workload's namespace.
    Workload WorkloadScalingPolicyAssignmentRuleRuleWorkload
    Allows assigning a scaling policy based on the workload's metadata.
    namespace object
    Allows assigning a scaling policy based on the workload's namespace.
    workload object
    Allows assigning a scaling policy based on the workload's metadata.
    namespace WorkloadScalingPolicyAssignmentRuleRuleNamespace
    Allows assigning a scaling policy based on the workload's namespace.
    workload WorkloadScalingPolicyAssignmentRuleRuleWorkload
    Allows assigning a scaling policy based on the workload's metadata.
    namespace workloadWorkloadScalingPolicyAssignmentRuleRuleNamespace
    Allows assigning a scaling policy based on the workload's namespace.
    workload workloadWorkloadScalingPolicyAssignmentRuleRuleWorkload
    Allows assigning a scaling policy based on the workload's metadata.
    namespace workload.WorkloadScalingPolicyAssignmentRuleRuleNamespace
    Allows assigning a scaling policy based on the workload's namespace.
    workload workload.WorkloadScalingPolicyAssignmentRuleRuleWorkload
    Allows assigning a scaling policy based on the workload's metadata.
    namespace Property Map
    Allows assigning a scaling policy based on the workload's namespace.
    workload Property Map
    Allows assigning a scaling policy based on the workload's metadata.

    WorkloadScalingPolicyAssignmentRuleRuleNamespace, WorkloadScalingPolicyAssignmentRuleRuleNamespaceArgs

    LabelsExpressions List<Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpression>
    Defines matching by label selector requirements.
    Names List<string>
    Defines matching by namespace names.
    LabelsExpressions WorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpression
    Defines matching by label selector requirements.
    Names []string
    Defines matching by namespace names.
    labels_expressions list(object)
    Defines matching by label selector requirements.
    names list(string)
    Defines matching by namespace names.
    labelsExpressions List<WorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpression>
    Defines matching by label selector requirements.
    names List<String>
    Defines matching by namespace names.
    labelsExpressions workloadWorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpression[]
    Defines matching by label selector requirements.
    names string[]
    Defines matching by namespace names.
    labels_expressions Sequence[workload.WorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpression]
    Defines matching by label selector requirements.
    names Sequence[str]
    Defines matching by namespace names.
    labelsExpressions List<Property Map>
    Defines matching by label selector requirements.
    names List<String>
    Defines matching by namespace names.

    WorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpression, WorkloadScalingPolicyAssignmentRuleRuleNamespaceLabelsExpressionArgs

    Operator string
    The operator to use for matching the label.
    Key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    Values List<string>
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    Operator string
    The operator to use for matching the label.
    Key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    Values []string
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator string
    The operator to use for matching the label.
    key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values list(string)
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator String
    The operator to use for matching the label.
    key String
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values List<String>
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator string
    The operator to use for matching the label.
    key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values string[]
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator str
    The operator to use for matching the label.
    key str
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values Sequence[str]
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator String
    The operator to use for matching the label.
    key String
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values List<String>
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.

    WorkloadScalingPolicyAssignmentRuleRuleWorkload, WorkloadScalingPolicyAssignmentRuleRuleWorkloadArgs

    Gvks List<string>
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    LabelsExpressions List<Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpression>
    Defines matching by label selector requirements.
    Gvks []string
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    LabelsExpressions WorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpression
    Defines matching by label selector requirements.
    gvks list(string)
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    labels_expressions list(object)
    Defines matching by label selector requirements.
    gvks List<String>
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    labelsExpressions List<WorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpression>
    Defines matching by label selector requirements.
    gvks string[]
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    labelsExpressions workloadWorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpression[]
    Defines matching by label selector requirements.
    gvks Sequence[str]
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    labels_expressions Sequence[workload.WorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpression]
    Defines matching by label selector requirements.
    gvks List<String>
    Group, version, and kind for Kubernetes resources. Format: kind[.version][.group]. It can be either:

    • only kind, e.g. "Deployment"
    • group and kind: e.g."Deployment.apps"
    • group, version and kind: e.g."Deployment.v1.apps"
    labelsExpressions List<Property Map>
    Defines matching by label selector requirements.

    WorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpression, WorkloadScalingPolicyAssignmentRuleRuleWorkloadLabelsExpressionArgs

    Operator string
    The operator to use for matching the label.
    Key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    Values List<string>
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    Operator string
    The operator to use for matching the label.
    Key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    Values []string
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator string
    The operator to use for matching the label.
    key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values list(string)
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator String
    The operator to use for matching the label.
    key String
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values List<String>
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator string
    The operator to use for matching the label.
    key string
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values string[]
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator str
    The operator to use for matching the label.
    key str
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values Sequence[str]
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.
    operator String
    The operator to use for matching the label.
    key String
    The label key to match. Required for all operators except Regex and Contains. If not specified, it will search through all labels.
    values List<String>
    A list of values to match against the label key. It is required for In, NotIn, Regex, and Contains operators.

    WorkloadScalingPolicyConfidence, WorkloadScalingPolicyConfidenceArgs

    Threshold double
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).
    Threshold float64
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).
    threshold number
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).
    threshold Double
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).
    threshold number
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).
    threshold float
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).
    threshold Number
    Defines the confidence threshold for applying recommendations. The smaller number indicates that we require fewer metrics data points to apply recommendations - changing this value can cause applying less precise recommendations. Do not change the default unless you want to optimize with fewer data points (e.g., short-lived workloads).

    WorkloadScalingPolicyCpu, WorkloadScalingPolicyCpuArgs

    ApplyThreshold double
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    ApplyThresholdStrategy Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyCpuApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    Args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    Constraints Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyCpuConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    Function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    Limit Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyCpuLimit
    Resource limit settings
    LookBackPeriodSeconds int
    The look back period in seconds for the recommendation.
    ManagementOption string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    Max double
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    Min double
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    Overhead double
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    ApplyThreshold float64
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    ApplyThresholdStrategy WorkloadScalingPolicyCpuApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    Args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    Constraints WorkloadScalingPolicyCpuConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    Function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    Limit WorkloadScalingPolicyCpuLimit
    Resource limit settings
    LookBackPeriodSeconds int
    The look back period in seconds for the recommendation.
    ManagementOption string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    Max float64
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    Min float64
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    Overhead float64
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    apply_threshold number
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    apply_threshold_strategy object
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints object
    Defines min/max bounds for the recommendation using constraint strategies.
    function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit object
    Resource limit settings
    look_back_period_seconds number
    The look back period in seconds for the recommendation.
    management_option string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max number
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min number
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead number
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    applyThreshold Double
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    applyThresholdStrategy WorkloadScalingPolicyCpuApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args String
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints WorkloadScalingPolicyCpuConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    function String
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit WorkloadScalingPolicyCpuLimit
    Resource limit settings
    lookBackPeriodSeconds Integer
    The look back period in seconds for the recommendation.
    managementOption String
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max Double
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min Double
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead Double
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    applyThreshold number
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    applyThresholdStrategy workloadWorkloadScalingPolicyCpuApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints workloadWorkloadScalingPolicyCpuConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit workloadWorkloadScalingPolicyCpuLimit
    Resource limit settings
    lookBackPeriodSeconds number
    The look back period in seconds for the recommendation.
    managementOption string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max number
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min number
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead number
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    apply_threshold float
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    apply_threshold_strategy workload.WorkloadScalingPolicyCpuApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args str
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints workload.WorkloadScalingPolicyCpuConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    function str
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit workload.WorkloadScalingPolicyCpuLimit
    Resource limit settings
    look_back_period_seconds int
    The look back period in seconds for the recommendation.
    management_option str
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max float
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min float
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead float
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    applyThreshold Number
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    applyThresholdStrategy Property Map
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args String
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints Property Map
    Defines min/max bounds for the recommendation using constraint strategies.
    function String
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit Property Map
    Resource limit settings
    lookBackPeriodSeconds Number
    The look back period in seconds for the recommendation.
    managementOption String
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max Number
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min Number
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead Number
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation

    WorkloadScalingPolicyCpuApplyThresholdStrategy, WorkloadScalingPolicyCpuApplyThresholdStrategyArgs

    Type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    Denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Exponent double
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    Numerator double
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Percentage double
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    Type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    Denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Exponent float64
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    Numerator float64
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Percentage float64
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent number
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator number
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage number
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type String
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator String
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent Double
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator Double
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage Double
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent number
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator number
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage number
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type str
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator str
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent float
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator float
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage float
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type String
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator String
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent Number
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator Number
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage Number
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.

    WorkloadScalingPolicyCpuConstraints, WorkloadScalingPolicyCpuConstraintsArgs

    WorkloadScalingPolicyCpuConstraintsMax, WorkloadScalingPolicyCpuConstraintsMaxArgs

    Constant double
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal double
    Bound as a percentage of the original pod-spec request.
    Constant float64
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal float64
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original number
    Bound as a percentage of the original pod-spec request.
    constant Double
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Double
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal number
    Bound as a percentage of the original pod-spec request.
    constant float
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original float
    Bound as a percentage of the original pod-spec request.
    constant Number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Number
    Bound as a percentage of the original pod-spec request.

    WorkloadScalingPolicyCpuConstraintsMin, WorkloadScalingPolicyCpuConstraintsMinArgs

    Constant double
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal double
    Bound as a percentage of the original pod-spec request.
    Constant float64
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal float64
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original number
    Bound as a percentage of the original pod-spec request.
    constant Double
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Double
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal number
    Bound as a percentage of the original pod-spec request.
    constant float
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original float
    Bound as a percentage of the original pod-spec request.
    constant Number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Number
    Bound as a percentage of the original pod-spec request.

    WorkloadScalingPolicyCpuLimit, WorkloadScalingPolicyCpuLimitArgs

    Type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    Multiplier double
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    OnlyIfOriginalExist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    OnlyIfOriginalLower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    Type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    Multiplier float64
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    OnlyIfOriginalExist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    OnlyIfOriginalLower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier number
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    only_if_original_exist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    only_if_original_lower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type String
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier Double
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    onlyIfOriginalExist Boolean
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    onlyIfOriginalLower Boolean
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier number
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    onlyIfOriginalExist boolean
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    onlyIfOriginalLower boolean
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type str
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier float
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    only_if_original_exist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    only_if_original_lower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type String
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier Number
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    onlyIfOriginalExist Boolean
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    onlyIfOriginalLower Boolean
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.

    WorkloadScalingPolicyDownscaling, WorkloadScalingPolicyDownscalingArgs

    ApplyType string
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    ApplyType string
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    apply_type string
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    applyType String
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    applyType string
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    apply_type str
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    applyType String
    Defines the apply type to be used when downscaling. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)

    WorkloadScalingPolicyHpaConverter, WorkloadScalingPolicyHpaConverterArgs

    Type string
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.
    Type string
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.
    type string
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.
    type String
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.
    type string
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.
    type str
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.
    type String
    HPA converter type. AVERAGE_VALUE_FROM_ORIGINAL_REQUESTS converts HPA utilization (%) targets to AverageValue using workload container requests.

    WorkloadScalingPolicyJvm, WorkloadScalingPolicyJvmArgs

    AutoInstrument bool
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    Memory Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyJvmMemory
    JVM memory optimization settings.
    AutoInstrument bool
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    Memory WorkloadScalingPolicyJvmMemory
    JVM memory optimization settings.
    auto_instrument bool
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    memory object
    JVM memory optimization settings.
    autoInstrument Boolean
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    memory WorkloadScalingPolicyJvmMemory
    JVM memory optimization settings.
    autoInstrument boolean
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    memory workloadWorkloadScalingPolicyJvmMemory
    JVM memory optimization settings.
    auto_instrument bool
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    memory workload.WorkloadScalingPolicyJvmMemory
    JVM memory optimization settings.
    autoInstrument Boolean
    When true, JMX exporter will be automatically injected into pods where JVM runtime is detected.
    memory Property Map
    JVM memory optimization settings.

    WorkloadScalingPolicyJvmMemory, WorkloadScalingPolicyJvmMemoryArgs

    Optimization bool
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.
    Optimization bool
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.
    optimization bool
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.
    optimization Boolean
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.
    optimization boolean
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.
    optimization bool
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.
    optimization Boolean
    Defines whether JVM memory optimization is enabled. When enabled, JVM heap size will be adjusted based on JVM metrics, if available.

    WorkloadScalingPolicyMemory, WorkloadScalingPolicyMemoryArgs

    ApplyThreshold double
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    ApplyThresholdStrategy Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemoryApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    Args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    Constraints Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemoryConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    Function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    Limit Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyMemoryLimit
    Resource limit settings
    LookBackPeriodSeconds int
    The look back period in seconds for the recommendation.
    ManagementOption string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    Max double
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    Min double
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    Overhead double
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    ApplyThreshold float64
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    ApplyThresholdStrategy WorkloadScalingPolicyMemoryApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    Args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    Constraints WorkloadScalingPolicyMemoryConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    Function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    Limit WorkloadScalingPolicyMemoryLimit
    Resource limit settings
    LookBackPeriodSeconds int
    The look back period in seconds for the recommendation.
    ManagementOption string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    Max float64
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    Min float64
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    Overhead float64
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    apply_threshold number
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    apply_threshold_strategy object
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints object
    Defines min/max bounds for the recommendation using constraint strategies.
    function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit object
    Resource limit settings
    look_back_period_seconds number
    The look back period in seconds for the recommendation.
    management_option string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max number
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min number
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead number
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    applyThreshold Double
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    applyThresholdStrategy WorkloadScalingPolicyMemoryApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args String
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints WorkloadScalingPolicyMemoryConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    function String
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit WorkloadScalingPolicyMemoryLimit
    Resource limit settings
    lookBackPeriodSeconds Integer
    The look back period in seconds for the recommendation.
    managementOption String
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max Double
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min Double
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead Double
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    applyThreshold number
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    applyThresholdStrategy workloadWorkloadScalingPolicyMemoryApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args string
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints workloadWorkloadScalingPolicyMemoryConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    function string
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit workloadWorkloadScalingPolicyMemoryLimit
    Resource limit settings
    lookBackPeriodSeconds number
    The look back period in seconds for the recommendation.
    managementOption string
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max number
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min number
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead number
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    apply_threshold float
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    apply_threshold_strategy workload.WorkloadScalingPolicyMemoryApplyThresholdStrategy
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args str
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints workload.WorkloadScalingPolicyMemoryConstraints
    Defines min/max bounds for the recommendation using constraint strategies.
    function str
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit workload.WorkloadScalingPolicyMemoryLimit
    Resource limit settings
    look_back_period_seconds int
    The look back period in seconds for the recommendation.
    management_option str
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max float
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min float
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead float
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation
    applyThreshold Number
    The threshold of when to apply the recommendation. Recommendation will be applied when diff of current requests and new recommendation is greater than set value

    Deprecated: Use applyThresholdStrategy instead

    applyThresholdStrategy Property Map
    Resource apply threshold strategy settings. The default strategy is PERCENTAGE with percentage value set to 0.1.
    args String
    The arguments for the function - i.e. for QUANTILE this should be a [0, 1] float. MAX doesn't accept any args
    constraints Property Map
    Defines min/max bounds for the recommendation using constraint strategies.
    function String
    The function used to calculate the resource recommendation. Supported values: QUANTILE, MAX
    limit Property Map
    Resource limit settings
    lookBackPeriodSeconds Number
    The look back period in seconds for the recommendation.
    managementOption String
    Disables management for a single resource when set to READ_ONLY. The resource will use its original workload template requests and limits. Supported value: READ_ONLY. Minimum required workload-autoscaler version: v0.23.1.
    max Number
    Max values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { max { constant = X } } instead

    min Number
    Min values for the recommendation, applies to every container. For memory - this is in MiB, for CPU - this is in cores.

    Deprecated: Use constraints { min { constant = X } } instead

    overhead Number
    Overhead for the recommendation, e.g. 0.1 will result in 10% higher recommendation

    WorkloadScalingPolicyMemoryApplyThresholdStrategy, WorkloadScalingPolicyMemoryApplyThresholdStrategyArgs

    Type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    Denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Exponent double
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    Numerator double
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Percentage double
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    Type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    Denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Exponent float64
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    Numerator float64
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    Percentage float64
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent number
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator number
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage number
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type String
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator String
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent Double
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator Double
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage Double
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type string
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator string
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent number
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator number
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage number
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type str
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator str
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent float
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator float
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage float
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.
    type String
    Defines apply theshold strategy type. - PERCENTAGE - recommendation will be applied when diff of current requests and new recommendation is greater than set value - DEFAULT_ADAPTIVE - will pick larger threshold percentage for small workloads and smaller percentage for large workloads. - CUSTOM_ADAPTIVE - works in same way as DEFAULT_ADAPTIVE, but it allows to tweak parameters of adaptive threshold formula: percentage = numerator/(currentRequest + denominator)^exponent. This strategy is for advance use cases, we recommend to use DEFAULT_ADAPTIVE strategy.
    denominator String
    If denominator is close or equal to 0, the threshold will be much bigger for small values.For example when numerator, exponent is 1 and denominator is 0 the threshold for 0.5 req. CPU will be 200%.It must be defined for the CUSTOM_ADAPTIVE strategy.
    exponent Number
    The exponent changes how fast the curve is going down. The smaller value will cause that we won’t pick extremely small number for big resources, for example: - if numerator is 0, denominator is 1, and exponent is 1, for 50 CPU we will pick 2% threshold - if numerator is 0, denominator is 1, and exponent is 0.8, for 50 CPU we will pick 4.3% threshold It must be defined for the CUSTOM_ADAPTIVE strategy.
    numerator Number
    The numerator affects vertical stretch of function used in adaptive threshold - smaller number will create smaller threshold.It must be defined for the CUSTOM_ADAPTIVE strategy.
    percentage Number
    Percentage of a how much difference should there be between the current pod requests and the new recommendation. It must be defined for the PERCENTAGE strategy.

    WorkloadScalingPolicyMemoryConstraints, WorkloadScalingPolicyMemoryConstraintsArgs

    WorkloadScalingPolicyMemoryConstraintsMax, WorkloadScalingPolicyMemoryConstraintsMaxArgs

    Constant double
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal double
    Bound as a percentage of the original pod-spec request.
    Constant float64
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal float64
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original number
    Bound as a percentage of the original pod-spec request.
    constant Double
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Double
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal number
    Bound as a percentage of the original pod-spec request.
    constant float
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original float
    Bound as a percentage of the original pod-spec request.
    constant Number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Number
    Bound as a percentage of the original pod-spec request.

    WorkloadScalingPolicyMemoryConstraintsMin, WorkloadScalingPolicyMemoryConstraintsMinArgs

    Constant double
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal double
    Bound as a percentage of the original pod-spec request.
    Constant float64
    Fixed bound value. For memory - MiB, for CPU - cores.
    PercentageOfOriginal float64
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original number
    Bound as a percentage of the original pod-spec request.
    constant Double
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Double
    Bound as a percentage of the original pod-spec request.
    constant number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal number
    Bound as a percentage of the original pod-spec request.
    constant float
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentage_of_original float
    Bound as a percentage of the original pod-spec request.
    constant Number
    Fixed bound value. For memory - MiB, for CPU - cores.
    percentageOfOriginal Number
    Bound as a percentage of the original pod-spec request.

    WorkloadScalingPolicyMemoryEvent, WorkloadScalingPolicyMemoryEventArgs

    ApplyType string
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    ApplyType string
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    apply_type string
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    applyType String
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    applyType string
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    apply_type str
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)
    applyType String
    Defines the apply type to be used when applying recommendation for memory related event. - IMMEDIATE - pods are restarted immediately when new recommendation is generated. - DEFERRED - pods are not restarted and recommendation values are applied during natural restarts only (new deployment, etc.)

    WorkloadScalingPolicyMemoryLimit, WorkloadScalingPolicyMemoryLimitArgs

    Type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    Multiplier double
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    OnlyIfOriginalExist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    OnlyIfOriginalLower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    Type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    Multiplier float64
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    OnlyIfOriginalExist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    OnlyIfOriginalLower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier number
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    only_if_original_exist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    only_if_original_lower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type String
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier Double
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    onlyIfOriginalExist Boolean
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    onlyIfOriginalLower Boolean
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type string
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier number
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    onlyIfOriginalExist boolean
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    onlyIfOriginalLower boolean
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type str
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier float
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    only_if_original_exist bool
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    only_if_original_lower bool
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.
    type String
    Defines limit strategy type. - NO_LIMIT - removes the resource limit even if it was specified in the workload spec. - KEEP_LIMITS - keep existing resource limits. While limits provide stability predictability, they may restrict workloads that need to temporarily burst beyond their allocation. - MULTIPLIER - used to calculate the resource limit. The final value is determined by multiplying the resource request by the specified factor. - MAINTAIN_RATIO - maintains the original ratio between requests and limits.
    multiplier Number
    Multiplier used to calculate the resource limit. It must be defined for the MULTIPLIER strategy.
    onlyIfOriginalExist Boolean
    When set to true, limits will only be set if the workload originally had limits defined in its manifest. If the original workload has no limits specified, no limits will be added.

    This flag allows conditional limit management based on the original workload configuration.

    Only applicable when the type is set to multiplier.
    onlyIfOriginalLower Boolean
    When set to true, limits will only be updated if the original limits are lower than the calculated value (requests × multiplier). If the original limits are already higher than the calculated value, they remain unchanged.

    This flag prevents reducing existing limits and ensures limits only increase when beneficial.

Only applicable when the type is set to multiplier.

    WorkloadScalingPolicyPredictiveScaling, WorkloadScalingPolicyPredictiveScalingArgs

    Cpu Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyPredictiveScalingCpu
    Defines predictive scaling resource configuration.
    Cpu WorkloadScalingPolicyPredictiveScalingCpu
    Defines predictive scaling resource configuration.
    cpu object
    Defines predictive scaling resource configuration.
    cpu WorkloadScalingPolicyPredictiveScalingCpu
    Defines predictive scaling resource configuration.
    cpu workloadWorkloadScalingPolicyPredictiveScalingCpu
    Defines predictive scaling resource configuration.
    cpu workload.WorkloadScalingPolicyPredictiveScalingCpu
    Defines predictive scaling resource configuration.
    cpu Property Map
    Defines predictive scaling resource configuration.

    WorkloadScalingPolicyPredictiveScalingCpu, WorkloadScalingPolicyPredictiveScalingCpuArgs

    Enabled bool
    Defines if predictive scaling is enabled for resource.
    Enabled bool
    Defines if predictive scaling is enabled for resource.
    enabled bool
    Defines if predictive scaling is enabled for resource.
    enabled Boolean
    Defines if predictive scaling is enabled for resource.
    enabled boolean
    Defines if predictive scaling is enabled for resource.
    enabled bool
    Defines if predictive scaling is enabled for resource.
    enabled Boolean
    Defines if predictive scaling is enabled for resource.

    WorkloadScalingPolicyRolloutBehavior, WorkloadScalingPolicyRolloutBehaviorArgs

    DelaySeconds int
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    PreferOneByOne bool
    Defines if pods should be restarted one by one to avoid service disruption.
    Type string
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.
    DelaySeconds int
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    PreferOneByOne bool
    Defines if pods should be restarted one by one to avoid service disruption.
    Type string
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.
    delay_seconds number
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    prefer_one_by_one bool
    Defines if pods should be restarted one by one to avoid service disruption.
    type string
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.
    delaySeconds Integer
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    preferOneByOne Boolean
    Defines if pods should be restarted one by one to avoid service disruption.
    type String
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.
    delaySeconds number
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    preferOneByOne boolean
    Defines if pods should be restarted one by one to avoid service disruption.
    type string
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.
    delay_seconds int
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    prefer_one_by_one bool
    Defines if pods should be restarted one by one to avoid service disruption.
    type str
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.
    delaySeconds Number
    Number of seconds to delay before applying the recommendation rollout. Must be between 0 and 3600.
    preferOneByOne Boolean
    Defines if pods should be restarted one by one to avoid service disruption.
    type String
    Defines the rollout type to be used when applying recommendations. - NO_DISRUPTION - pods are restarted without causing service disruption. - UNSPECIFIED - rollout type is not specified.

    WorkloadScalingPolicyStartup, WorkloadScalingPolicyStartupArgs

    PeriodSeconds int
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    TwoPhaseRecommendations Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyStartupTwoPhaseRecommendations
    Defines two-phase recommendations settings for the startup period.
    PeriodSeconds int
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    TwoPhaseRecommendations WorkloadScalingPolicyStartupTwoPhaseRecommendations
    Defines two-phase recommendations settings for the startup period.
    period_seconds number
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    two_phase_recommendations object
    Defines two-phase recommendations settings for the startup period.
    periodSeconds Integer
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    twoPhaseRecommendations WorkloadScalingPolicyStartupTwoPhaseRecommendations
    Defines two-phase recommendations settings for the startup period.
    periodSeconds number
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    twoPhaseRecommendations workloadWorkloadScalingPolicyStartupTwoPhaseRecommendations
    Defines two-phase recommendations settings for the startup period.
    period_seconds int
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    two_phase_recommendations workload.WorkloadScalingPolicyStartupTwoPhaseRecommendations
    Defines two-phase recommendations settings for the startup period.
    periodSeconds Number
    Defines the duration (in seconds) during which elevated resource usage is expected at startup. When set, recommendations will be adjusted to disregard resource spikes within this period. If not specified, the workload will receive standard recommendations without startup considerations.
    twoPhaseRecommendations Property Map
    Defines two-phase recommendations settings for the startup period.

    WorkloadScalingPolicyStartupTwoPhaseRecommendations, WorkloadScalingPolicyStartupTwoPhaseRecommendationsArgs

    Enabled bool
    Defines whether two-phase recommendations are enabled during startup.
    RequestsOnStartup Pulumi.CastAI.Workload.Inputs.WorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartup
    Defines the resource requests to use during startup.
    Enabled bool
    Defines whether two-phase recommendations are enabled during startup.
    RequestsOnStartup WorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartup
    Defines the resource requests to use during startup.
    enabled bool
    Defines whether two-phase recommendations are enabled during startup.
    requests_on_startup object
    Defines the resource requests to use during startup.
    enabled Boolean
    Defines whether two-phase recommendations are enabled during startup.
    requestsOnStartup WorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartup
    Defines the resource requests to use during startup.
    enabled boolean
    Defines whether two-phase recommendations are enabled during startup.
    requestsOnStartup workloadWorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartup
    Defines the resource requests to use during startup.
    enabled bool
    Defines whether two-phase recommendations are enabled during startup.
    requests_on_startup workload.WorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartup
    Defines the resource requests to use during startup.
    enabled Boolean
    Defines whether two-phase recommendations are enabled during startup.
    requestsOnStartup Property Map
    Defines the resource requests to use during startup.

    WorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartup, WorkloadScalingPolicyStartupTwoPhaseRecommendationsRequestsOnStartupArgs

    CpuCores double
    CPU cores to request during startup.
    MemoryGib double
    Memory in GiB to request during startup.
    CpuCores float64
    CPU cores to request during startup.
    MemoryGib float64
    Memory in GiB to request during startup.
    cpu_cores number
    CPU cores to request during startup.
    memory_gib number
    Memory in GiB to request during startup.
    cpuCores Double
    CPU cores to request during startup.
    memoryGib Double
    Memory in GiB to request during startup.
    cpuCores number
    CPU cores to request during startup.
    memoryGib number
    Memory in GiB to request during startup.
    cpu_cores float
    CPU cores to request during startup.
    memory_gib float
    Memory in GiB to request during startup.
    cpuCores Number
    CPU cores to request during startup.
    memoryGib Number
    Memory in GiB to request during startup.

    Package Details

    Repository
    castai castai/pulumi-castai
    License
    Apache-2.0
    Notes
    This Pulumi package is based on the castai Terraform Provider.
    castai logo
    Viewing docs for CAST AI v8.60.0
    published on Wednesday, Aug 19, 2026 by CAST AI
    Schema (JSON)
    castai/pulumi-castai

      Try Pulumi Cloud free.
      Your team will thank you.

      Start free trial