v1beta2

package
v0.0.0-...-1dfa148 Latest Latest
Warning

This package is not in the latest version of its module.

Go to latest
Published: Mar 12, 2024 License: Apache-2.0 Imports: 5 Imported by: 0

Documentation

Overview

Package v1beta2 is the v1beta2 version of the API. +groupName=sparkoperator.k8s.io +versionName=v1beta2

Index

Constants

View Source
const Version = "v1beta2"

Variables

View Source
var (
	SchemeBuilder = runtime.NewSchemeBuilder(addKnownTypes)
	AddToScheme   = SchemeBuilder.AddToScheme
)
View Source
var SchemeGroupVersion = schema.GroupVersion{Group: sparkoperator.GroupName, Version: Version}

SchemeGroupVersion is the group version used to register these objects.

Functions

func Resource

func Resource(resource string) schema.GroupResource

Resource takes an unqualified resource and returns a Group-qualified GroupResource.

func SetSparkApplicationDefaults

func SetSparkApplicationDefaults(app *SparkApplication)

SetSparkApplicationDefaults sets default values for certain fields of a SparkApplication.

Types

type ApplicationState

type ApplicationState struct {
	State        ApplicationStateType `json:"state"`
	ErrorMessage string               `json:"errorMessage,omitempty"`
}

ApplicationState tells the current state of the application and an error message in case of failures.

func (*ApplicationState) DeepCopy

func (in *ApplicationState) DeepCopy() *ApplicationState

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ApplicationState.

func (*ApplicationState) DeepCopyInto

func (in *ApplicationState) DeepCopyInto(out *ApplicationState)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type ApplicationStateType

type ApplicationStateType string

ApplicationStateType represents the type of the current state of an application.

const (
	NewState              ApplicationStateType = ""
	SubmittedState        ApplicationStateType = "SUBMITTED"
	RunningState          ApplicationStateType = "RUNNING"
	CompletedState        ApplicationStateType = "COMPLETED"
	FailedState           ApplicationStateType = "FAILED"
	FailedSubmissionState ApplicationStateType = "SUBMISSION_FAILED"
	PendingRerunState     ApplicationStateType = "PENDING_RERUN"
	InvalidatingState     ApplicationStateType = "INVALIDATING"
	SucceedingState       ApplicationStateType = "SUCCEEDING"
	FailingState          ApplicationStateType = "FAILING"
	UnknownState          ApplicationStateType = "UNKNOWN"
	KilledState           ApplicationStateType = "KILLED"
)

Different states an application may have.

type BatchSchedulerConfiguration

type BatchSchedulerConfiguration struct {
	// Queue stands for the resource queue which the application belongs to, it's being used in Volcano batch scheduler.
	// +optional
	Queue *string `json:"queue,omitempty"`
	// PriorityClassName stands for the name of k8s PriorityClass resource, it's being used in Volcano batch scheduler.
	// +optional
	PriorityClassName *string `json:"priorityClassName,omitempty"`
}

BatchSchedulerConfiguration used to configure how to batch scheduling Spark Application

func (*BatchSchedulerConfiguration) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new BatchSchedulerConfiguration.

func (*BatchSchedulerConfiguration) DeepCopyInto

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type ConcurrencyPolicy

type ConcurrencyPolicy string
const (
	// ConcurrencyAllow allows SparkApplications to run concurrently.
	ConcurrencyAllow ConcurrencyPolicy = "Allow"
	// ConcurrencyForbid forbids concurrent runs of SparkApplications, skipping the next run if the previous
	// one hasn't finished yet.
	ConcurrencyForbid ConcurrencyPolicy = "Forbid"
	// ConcurrencyReplace kills the currently running SparkApplication instance and replaces it with a new one.
	ConcurrencyReplace ConcurrencyPolicy = "Replace"
)

type Dependencies

type Dependencies struct {
	// Jars is a list of JAR files the Spark application depends on.
	// +optional
	Jars []string `json:"jars,omitempty"`
	// Files is a list of files the Spark application depends on.
	// +optional
	Files []string `json:"files,omitempty"`
	// PyFiles is a list of Python files the Spark application depends on.
	// +optional
	PyFiles []string `json:"pyFiles,omitempty"`
}

Dependencies specifies all possible types of dependencies of a Spark application.

func (*Dependencies) DeepCopy

func (in *Dependencies) DeepCopy() *Dependencies

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new Dependencies.

func (*Dependencies) DeepCopyInto

func (in *Dependencies) DeepCopyInto(out *Dependencies)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type DeployMode

type DeployMode string

DeployMode describes the type of deployment of a Spark application.

const (
	ClusterMode         DeployMode = "cluster"
	ClientMode          DeployMode = "client"
	InClusterClientMode DeployMode = "in-cluster-client"
)

Different types of deployments.

type DriverInfo

type DriverInfo struct {
	WebUIServiceName string `json:"webUIServiceName,omitempty"`
	// UI Details for the UI created via ClusterIP service accessible from within the cluster.
	WebUIPort    int32  `json:"webUIPort,omitempty"`
	WebUIAddress string `json:"webUIAddress,omitempty"`
	// Ingress Details if an ingress for the UI was created.
	WebUIIngressName    string `json:"webUIIngressName,omitempty"`
	WebUIIngressAddress string `json:"webUIIngressAddress,omitempty"`
	PodName             string `json:"podName,omitempty"`
	PodState            string `json:"podState,omitempty" `
	// +nullable
	CreationTimestamp metav1.Time `json:"creationTimestamp,omitempty"`
	// +nullable
	TerminationTime metav1.Time `json:"terminationTime,omitempty"`
}

DriverInfo captures information about the driver.

func (*DriverInfo) DeepCopy

func (in *DriverInfo) DeepCopy() *DriverInfo

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new DriverInfo.

func (*DriverInfo) DeepCopyInto

func (in *DriverInfo) DeepCopyInto(out *DriverInfo)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type DriverSpec

type DriverSpec struct {
	SparkPodSpec `json:",inline"`
	// PodName is the name of the driver pod that the user creates. This is used for the
	// in-cluster client mode in which the user creates a client pod where the driver of
	// the user application runs. It's an error to set this field if Mode is not
	// in-cluster-client.
	// +optional
	// +kubebuilder:validation:Pattern=[a-z0-9]([-a-z0-9]*[a-z0-9])?(\\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*
	PodName *string `json:"podName,omitempty"`
	// CoreRequest is the physical CPU core request for the driver.
	// Maps to `spark.kubernetes.driver.request.cores` that is available since Spark 3.0.
	// +optional
	CoreRequest *string `json:"coreRequest,omitempty"`
	// ServiceAccount is the name of the Kubernetes service account used by the driver pod
	// when requesting executor pods from the API server.
	// +optional
	ServiceAccount *string `json:"serviceAccount,omitempty"`
	// JavaOptions is a string of extra JVM options to pass to the driver. For instance,
	// GC settings or other logging.
	// +optional
	JavaOptions *string `json:"javaOptions,omitempty"`
	// Lifecycle for running preStop or postStart commands
	// +optional
	Lifecycle *apiv1.Lifecycle `json:"lifecycle,omitempty"`
}

DriverSpec is specification of the driver.

func (*DriverSpec) DeepCopy

func (in *DriverSpec) DeepCopy() *DriverSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new DriverSpec.

func (*DriverSpec) DeepCopyInto

func (in *DriverSpec) DeepCopyInto(out *DriverSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type ExecutorSpec

type ExecutorSpec struct {
	SparkPodSpec `json:",inline"`
	// Instances is the number of executor instances.
	// +optional
	// +kubebuilder:validation:Minimum=1
	Instances *int32 `json:"instances,omitempty"`
	// CoreRequest is the physical CPU core request for the executors.
	// Maps to `spark.kubernetes.executor.request.cores` that is available since Spark 2.4.
	// +optional
	CoreRequest *string `json:"coreRequest,omitempty"`
	// JavaOptions is a string of extra JVM options to pass to the executors. For instance,
	// GC settings or other logging.
	// +optional
	JavaOptions *string `json:"javaOptions,omitempty"`
	// DeleteOnTermination specify whether executor pods should be deleted in case of failure or normal termination.
	// Maps to `spark.kubernetes.executor.deleteOnTermination` that is available since Spark 3.0.
	// +optional
	DeleteOnTermination *bool `json:"deleteOnTermination,omitempty"`
}

ExecutorSpec is specification of the executor.

func (*ExecutorSpec) DeepCopy

func (in *ExecutorSpec) DeepCopy() *ExecutorSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExecutorSpec.

func (*ExecutorSpec) DeepCopyInto

func (in *ExecutorSpec) DeepCopyInto(out *ExecutorSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type ExecutorState

type ExecutorState string

ExecutorState tells the current state of an executor.

const (
	ExecutorPendingState   ExecutorState = "PENDING"
	ExecutorRunningState   ExecutorState = "RUNNING"
	ExecutorCompletedState ExecutorState = "COMPLETED"
	ExecutorFailedState    ExecutorState = "FAILED"
	ExecutorUnknownState   ExecutorState = "UNKNOWN"
)

Different states an executor may have.

type GPUSpec

type GPUSpec struct {
	// Name is GPU resource name, such as: nvidia.com/gpu or amd.com/gpu
	Name string `json:"name"`
	// Quantity is the number of GPUs to request for driver or executor.
	Quantity int64 `json:"quantity"`
}

func (*GPUSpec) DeepCopy

func (in *GPUSpec) DeepCopy() *GPUSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new GPUSpec.

func (*GPUSpec) DeepCopyInto

func (in *GPUSpec) DeepCopyInto(out *GPUSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type MonitoringSpec

type MonitoringSpec struct {
	// ExposeDriverMetrics specifies whether to expose metrics on the driver.
	ExposeDriverMetrics bool `json:"exposeDriverMetrics"`
	// ExposeExecutorMetrics specifies whether to expose metrics on the executors.
	ExposeExecutorMetrics bool `json:"exposeExecutorMetrics"`
	// MetricsProperties is the content of a custom metrics.properties for configuring the Spark metric system.
	// +optional
	// If not specified, the content in spark-docker/conf/metrics.properties will be used.
	MetricsProperties *string `json:"metricsProperties,omitempty"`
	// Prometheus is for configuring the Prometheus JMX exporter.
	// +optional
	Prometheus *PrometheusSpec `json:"prometheus,omitempty"`
}

MonitoringSpec defines the monitoring specification.

func (*MonitoringSpec) DeepCopy

func (in *MonitoringSpec) DeepCopy() *MonitoringSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MonitoringSpec.

func (*MonitoringSpec) DeepCopyInto

func (in *MonitoringSpec) DeepCopyInto(out *MonitoringSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type NameKey

type NameKey struct {
	Name string `json:"name"`
	Key  string `json:"key"`
}

NameKey represents the name and key of a SecretKeyRef.

func (*NameKey) DeepCopy

func (in *NameKey) DeepCopy() *NameKey

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new NameKey.

func (*NameKey) DeepCopyInto

func (in *NameKey) DeepCopyInto(out *NameKey)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type NamePath

type NamePath struct {
	Name string `json:"name"`
	Path string `json:"path"`
}

NamePath is a pair of a name and a path to which the named objects should be mounted to.

func (*NamePath) DeepCopy

func (in *NamePath) DeepCopy() *NamePath

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new NamePath.

func (*NamePath) DeepCopyInto

func (in *NamePath) DeepCopyInto(out *NamePath)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type PrometheusSpec

type PrometheusSpec struct {
	// JmxExporterJar is the path to the Prometheus JMX exporter jar in the container.
	JmxExporterJar string `json:"jmxExporterJar"`
	// Port is the port of the HTTP server run by the Prometheus JMX exporter.
	// If not specified, 8090 will be used as the default.
	// +kubebuilder:validation:Minimum=1024
	// +kubebuilder:validation:Maximum=49151
	// +optional
	Port *int32 `json:"port,omitempty"`
	// ConfigFile is the path to the custom Prometheus configuration file provided in the Spark image.
	// ConfigFile takes precedence over Configuration, which is shown below.
	// +optional
	ConfigFile *string `json:"configFile,omitempty"`
	// Configuration is the content of the Prometheus configuration needed by the Prometheus JMX exporter.
	// If not specified, the content in spark-docker/conf/prometheus.yaml will be used.
	// Configuration has no effect if ConfigFile is set.
	// +optional
	Configuration *string `json:"configuration,omitempty"`
}

PrometheusSpec defines the Prometheus specification when Prometheus is to be used for collecting and exposing metrics.

func (*PrometheusSpec) DeepCopy

func (in *PrometheusSpec) DeepCopy() *PrometheusSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PrometheusSpec.

func (*PrometheusSpec) DeepCopyInto

func (in *PrometheusSpec) DeepCopyInto(out *PrometheusSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type RestartPolicy

type RestartPolicy struct {
	// Type specifies the RestartPolicyType.
	// +kubebuilder:validation:Enum={Never,Always,OnFailure}
	Type RestartPolicyType `json:"type,omitempty"`

	// OnSubmissionFailureRetries is the number of times to retry submitting an application before giving up.
	// This is best effort and actual retry attempts can be >= the value specified due to caching.
	// These are required if RestartPolicy is OnFailure.
	// +kubebuilder:validation:Minimum=0
	// +optional
	OnSubmissionFailureRetries *int32 `json:"onSubmissionFailureRetries,omitempty"`

	// OnFailureRetries the number of times to retry running an application before giving up.
	// +kubebuilder:validation:Minimum=0
	// +optional
	OnFailureRetries *int32 `json:"onFailureRetries,omitempty"`

	// OnSubmissionFailureRetryInterval is the interval between retries on failed submissions.
	// Interval to wait between successive retries of a failed application.
	// +kubebuilder:validation:Minimum=1
	// +optional
	OnSubmissionFailureRetryInterval *int64 `json:"onSubmissionFailureRetryInterval,omitempty"`

	// OnFailureRetryInterval is the interval between retries on failed runs.
	// +kubebuilder:validation:Minimum=1
	// +optional
	OnFailureRetryInterval *int64 `json:"onFailureRetryInterval,omitempty"`
}

RestartPolicy is the policy of if and in which conditions the controller should restart a terminated application. This completely defines actions to be taken on any kind of Failures during an application run.

func (*RestartPolicy) DeepCopy

func (in *RestartPolicy) DeepCopy() *RestartPolicy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new RestartPolicy.

func (*RestartPolicy) DeepCopyInto

func (in *RestartPolicy) DeepCopyInto(out *RestartPolicy)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type RestartPolicyType

type RestartPolicyType string
const (
	Never     RestartPolicyType = "Never"
	OnFailure RestartPolicyType = "OnFailure"
	Always    RestartPolicyType = "Always"
)

type ScheduleState

type ScheduleState string
const (
	FailedValidationState ScheduleState = "FailedValidation"
	ScheduledState        ScheduleState = "Scheduled"
)

type ScheduledSparkApplication

type ScheduledSparkApplication struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata"`
	Spec              ScheduledSparkApplicationSpec   `json:"spec"`
	Status            ScheduledSparkApplicationStatus `json:"status,omitempty"`
}

func (*ScheduledSparkApplication) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ScheduledSparkApplication.

func (*ScheduledSparkApplication) DeepCopyInto

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*ScheduledSparkApplication) DeepCopyObject

func (in *ScheduledSparkApplication) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type ScheduledSparkApplicationList

type ScheduledSparkApplicationList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	Items           []ScheduledSparkApplication `json:"items,omitempty"`
}

ScheduledSparkApplicationList carries a list of ScheduledSparkApplication objects.

func (*ScheduledSparkApplicationList) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ScheduledSparkApplicationList.

func (*ScheduledSparkApplicationList) DeepCopyInto

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*ScheduledSparkApplicationList) DeepCopyObject

func (in *ScheduledSparkApplicationList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type ScheduledSparkApplicationSpec

type ScheduledSparkApplicationSpec struct {
	// Schedule is a cron schedule on which the application should run.
	Schedule string `json:"schedule"`
	// Template is a template from which SparkApplication instances can be created.
	Template SparkApplicationSpec `json:"template"`
	// Suspend is a flag telling the controller to suspend subsequent runs of the application if set to true.
	// +optional
	// Defaults to false.
	Suspend *bool `json:"suspend,omitempty"`
	// ConcurrencyPolicy is the policy governing concurrent SparkApplication runs.
	ConcurrencyPolicy ConcurrencyPolicy `json:"concurrencyPolicy,omitempty"`
	// SuccessfulRunHistoryLimit is the number of past successful runs of the application to keep.
	// +optional
	// Defaults to 1.
	SuccessfulRunHistoryLimit *int32 `json:"successfulRunHistoryLimit,omitempty"`
	// FailedRunHistoryLimit is the number of past failed runs of the application to keep.
	// +optional
	// Defaults to 1.
	FailedRunHistoryLimit *int32 `json:"failedRunHistoryLimit,omitempty"`
}

func (*ScheduledSparkApplicationSpec) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ScheduledSparkApplicationSpec.

func (*ScheduledSparkApplicationSpec) DeepCopyInto

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type ScheduledSparkApplicationStatus

type ScheduledSparkApplicationStatus struct {
	// LastRun is the time when the last run of the application started.
	// +nullable
	LastRun metav1.Time `json:"lastRun,omitempty"`
	// NextRun is the time when the next run of the application will start.
	// +nullable
	NextRun metav1.Time `json:"nextRun,omitempty"`
	// LastRunName is the name of the SparkApplication for the most recent run of the application.
	LastRunName string `json:"lastRunName,omitempty"`
	// PastSuccessfulRunNames keeps the names of SparkApplications for past successful runs.
	PastSuccessfulRunNames []string `json:"pastSuccessfulRunNames,omitempty"`
	// PastFailedRunNames keeps the names of SparkApplications for past failed runs.
	PastFailedRunNames []string `json:"pastFailedRunNames,omitempty"`
	// ScheduleState is the current scheduling state of the application.
	ScheduleState ScheduleState `json:"scheduleState,omitempty"`
	// Reason tells why the ScheduledSparkApplication is in the particular ScheduleState.
	Reason string `json:"reason,omitempty"`
}

func (*ScheduledSparkApplicationStatus) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ScheduledSparkApplicationStatus.

func (*ScheduledSparkApplicationStatus) DeepCopyInto

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type SecretInfo

type SecretInfo struct {
	Name string     `json:"name"`
	Path string     `json:"path"`
	Type SecretType `json:"secretType"`
}

SecretInfo captures information of a secret.

func (*SecretInfo) DeepCopy

func (in *SecretInfo) DeepCopy() *SecretInfo

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SecretInfo.

func (*SecretInfo) DeepCopyInto

func (in *SecretInfo) DeepCopyInto(out *SecretInfo)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type SecretType

type SecretType string

SecretType tells the type of a secret.

const (
	// GCPServiceAccountSecret is for secrets from a GCP service account Json key file that needs
	// the environment variable GOOGLE_APPLICATION_CREDENTIALS.
	GCPServiceAccountSecret SecretType = "GCPServiceAccount"
	// HadoopDelegationTokenSecret is for secrets from an Hadoop delegation token that needs the
	// environment variable HADOOP_TOKEN_FILE_LOCATION.
	HadoopDelegationTokenSecret SecretType = "HadoopDelegationToken"
	// GenericType is for secrets that needs no special handling.
	GenericType SecretType = "Generic"
)

An enumeration of secret types supported.

type SparkApplication

type SparkApplication struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata"`
	Spec              SparkApplicationSpec   `json:"spec"`
	Status            SparkApplicationStatus `json:"status,omitempty"`
}

SparkApplication represents a Spark application running on and using Kubernetes as a cluster manager.

func (*SparkApplication) DeepCopy

func (in *SparkApplication) DeepCopy() *SparkApplication

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SparkApplication.

func (*SparkApplication) DeepCopyInto

func (in *SparkApplication) DeepCopyInto(out *SparkApplication)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*SparkApplication) DeepCopyObject

func (in *SparkApplication) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

func (*SparkApplication) ExposeDriverMetrics

func (s *SparkApplication) ExposeDriverMetrics() bool

ExposeDriverMetrics returns if driver metrics should be exposed.

func (*SparkApplication) ExposeExecutorMetrics

func (s *SparkApplication) ExposeExecutorMetrics() bool

ExposeExecutorMetrics returns if executor metrics should be exposed.

func (*SparkApplication) HasPrometheusConfigFile

func (s *SparkApplication) HasPrometheusConfigFile() bool

HasPrometheusConfigFile returns if Prometheus monitoring uses a configruation file in the container.

func (*SparkApplication) PrometheusMonitoringEnabled

func (s *SparkApplication) PrometheusMonitoringEnabled() bool

PrometheusMonitoringEnabled returns if Prometheus monitoring is enabled or not.

type SparkApplicationList

type SparkApplicationList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	Items           []SparkApplication `json:"items,omitempty"`
}

SparkApplicationList carries a list of SparkApplication objects.

func (*SparkApplicationList) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SparkApplicationList.

func (*SparkApplicationList) DeepCopyInto

func (in *SparkApplicationList) DeepCopyInto(out *SparkApplicationList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*SparkApplicationList) DeepCopyObject

func (in *SparkApplicationList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type SparkApplicationSpec

type SparkApplicationSpec struct {
	// Type tells the type of the Spark application.
	// +kubebuilder:validation:Enum={Java,Python,Scala,R}
	Type SparkApplicationType `json:"type"`
	// SparkVersion is the version of Spark the application uses.
	SparkVersion string `json:"sparkVersion"`
	// Mode is the deployment mode of the Spark application.
	// +kubebuilder:validation:Enum={cluster,client}
	Mode DeployMode `json:"mode,omitempty"`
	// ProxyUser specifies the user to impersonate when submitting the application.
	// It maps to the command-line flag "--proxy-user" in spark-submit.
	// +optional
	ProxyUser *string `json:"proxyUser,omitempty"`
	// Image is the container image for the driver, executor, and init-container. Any custom container images for the
	// driver, executor, or init-container takes precedence over this.
	// +optional
	Image *string `json:"image,omitempty"`
	// ImagePullPolicy is the image pull policy for the driver, executor, and init-container.
	// +optional
	ImagePullPolicy *string `json:"imagePullPolicy,omitempty"`
	// ImagePullSecrets is the list of image-pull secrets.
	// +optional
	ImagePullSecrets []string `json:"imagePullSecrets,omitempty"`
	// MainClass is the fully-qualified main class of the Spark application.
	// This only applies to Java/Scala Spark applications.
	// +optional
	MainClass *string `json:"mainClass,omitempty"`
	// MainFile is the path to a bundled JAR, Python, or R file of the application.
	// +optional
	MainApplicationFile *string `json:"mainApplicationFile"`
	// Arguments is a list of arguments to be passed to the application.
	// +optional
	Arguments []string `json:"arguments,omitempty"`
	// SparkConf carries user-specified Spark configuration properties as they would use the  "--conf" option in
	// spark-submit.
	// +optional
	SparkConf map[string]string `json:"sparkConf,omitempty"`
	// HadoopConf carries user-specified Hadoop configuration properties as they would use the  the "--conf" option
	// in spark-submit.  The SparkApplication controller automatically adds prefix "spark.hadoop." to Hadoop
	// configuration properties.
	// +optional
	HadoopConf map[string]string `json:"hadoopConf,omitempty"`
	// SparkConfigMap carries the name of the ConfigMap containing Spark configuration files such as log4j.properties.
	// The controller will add environment variable SPARK_CONF_DIR to the path where the ConfigMap is mounted to.
	// +optional
	SparkConfigMap *string `json:"sparkConfigMap,omitempty"`
	// HadoopConfigMap carries the name of the ConfigMap containing Hadoop configuration files such as core-site.xml.
	// The controller will add environment variable HADOOP_CONF_DIR to the path where the ConfigMap is mounted to.
	// +optional
	HadoopConfigMap *string `json:"hadoopConfigMap,omitempty"`
	// Volumes is the list of Kubernetes volumes that can be mounted by the driver and/or executors.
	// +optional
	Volumes []apiv1.Volume `json:"volumes,omitempty"`
	// Driver is the driver specification.
	Driver DriverSpec `json:"driver"`
	// Executor is the executor specification.
	Executor ExecutorSpec `json:"executor"`
	// Deps captures all possible types of dependencies of a Spark application.
	// +optional
	Deps Dependencies `json:"deps,omitempty"`
	// RestartPolicy defines the policy on if and in which conditions the controller should restart an application.
	RestartPolicy RestartPolicy `json:"restartPolicy,omitempty"`
	// NodeSelector is the Kubernetes node selector to be added to the driver and executor pods.
	// This field is mutually exclusive with nodeSelector at podSpec level (driver or executor).
	// This field will be deprecated in future versions (at SparkApplicationSpec level).
	// +optional
	NodeSelector map[string]string `json:"nodeSelector,omitempty"`
	// FailureRetries is the number of times to retry a failed application before giving up.
	// This is best effort and actual retry attempts can be >= the value specified.
	// +optional
	FailureRetries *int32 `json:"failureRetries,omitempty"`
	// RetryInterval is the unit of intervals in seconds between submission retries.
	// +optional
	RetryInterval *int64 `json:"retryInterval,omitempty"`
	// This sets the major Python version of the docker
	// image used to run the driver and executor containers. Can either be 2 or 3, default 2.
	// +optional
	// +kubebuilder:validation:Enum={"2","3"}
	PythonVersion *string `json:"pythonVersion,omitempty"`
	// This sets the Memory Overhead Factor that will allocate memory to non-JVM memory.
	// For JVM-based jobs this value will default to 0.10, for non-JVM jobs 0.40. Value of this field will
	// be overridden by `Spec.Driver.MemoryOverhead` and `Spec.Executor.MemoryOverhead` if they are set.
	// +optional
	MemoryOverheadFactor *string `json:"memoryOverheadFactor,omitempty"`
	// Monitoring configures how monitoring is handled.
	// +optional
	Monitoring *MonitoringSpec `json:"monitoring,omitempty"`
	// BatchScheduler configures which batch scheduler will be used for scheduling
	// +optional
	BatchScheduler *string `json:"batchScheduler,omitempty"`
	// TimeToLiveSeconds defines the Time-To-Live (TTL) duration in seconds for this SparkAplication
	// after its termination.
	// The SparkApplication object will be garbage collected if the current time is more than the
	// TimeToLiveSeconds since its termination.
	// +optional
	TimeToLiveSeconds *int64 `json:"timeToLiveSeconds,omitempty"`
	// BatchSchedulerOptions provides fine-grained control on how to batch scheduling.
	// +optional
	BatchSchedulerOptions *BatchSchedulerConfiguration `json:"batchSchedulerOptions,omitempty"`
}

SparkApplicationSpec describes the specification of a Spark application using Kubernetes as a cluster manager. It carries every pieces of information a spark-submit command takes and recognizes.

func (*SparkApplicationSpec) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SparkApplicationSpec.

func (*SparkApplicationSpec) DeepCopyInto

func (in *SparkApplicationSpec) DeepCopyInto(out *SparkApplicationSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type SparkApplicationStatus

type SparkApplicationStatus struct {
	// SparkApplicationID is set by the spark-distribution(via spark.app.id config) on the driver and executor pods
	SparkApplicationID string `json:"sparkApplicationId,omitempty"`
	// SubmissionID is a unique ID of the current submission of the application.
	SubmissionID string `json:"submissionID,omitempty"`
	// LastSubmissionAttemptTime is the time for the last application submission attempt.
	// +nullable
	LastSubmissionAttemptTime metav1.Time `json:"lastSubmissionAttemptTime,omitempty"`
	// CompletionTime is the time when the application runs to completion if it does.
	// +nullable
	TerminationTime metav1.Time `json:"terminationTime,omitempty"`
	// DriverInfo has information about the driver.
	DriverInfo DriverInfo `json:"driverInfo"`
	// AppState tells the overall application state.
	AppState ApplicationState `json:"applicationState,omitempty"`
	// ExecutorState records the state of executors by executor Pod names.
	ExecutorState map[string]string `json:"executorState,omitempty"`
	// ExecutionAttempts is the total number of attempts to run a submitted application to completion.
	// Incremented upon each attempted run of the application and reset upon invalidation.
	ExecutionAttempts int32 `json:"executionAttempts,omitempty"`
	// SubmissionAttempts is the total number of attempts to submit an application to run.
	// Incremented upon each attempted submission of the application and reset upon invalidation and rerun.
	SubmissionAttempts int32 `json:"submissionAttempts,omitempty"`
}

SparkApplicationStatus describes the current status of a Spark application.

func (*SparkApplicationStatus) DeepCopy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SparkApplicationStatus.

func (*SparkApplicationStatus) DeepCopyInto

func (in *SparkApplicationStatus) DeepCopyInto(out *SparkApplicationStatus)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type SparkApplicationType

type SparkApplicationType string

SparkApplicationType describes the type of a Spark application.

const (
	JavaApplicationType   SparkApplicationType = "Java"
	ScalaApplicationType  SparkApplicationType = "Scala"
	PythonApplicationType SparkApplicationType = "Python"
	RApplicationType      SparkApplicationType = "R"
)

Different types of Spark applications.

type SparkPodSpec

type SparkPodSpec struct {
	// Cores maps to `spark.driver.cores` or `spark.executor.cores` for the driver and executors, respectively.
	// +optional
	// +kubebuilder:validation:Minimum=1
	Cores *int32 `json:"cores,omitempty"`
	// CoreLimit specifies a hard limit on CPU cores for the pod.
	// Optional
	CoreLimit *string `json:"coreLimit,omitempty"`
	// Memory is the amount of memory to request for the pod.
	// +optional
	Memory *string `json:"memory,omitempty"`
	// MemoryOverhead is the amount of off-heap memory to allocate in cluster mode, in MiB unless otherwise specified.
	// +optional
	MemoryOverhead *string `json:"memoryOverhead,omitempty"`
	// GPU specifies GPU requirement for the pod.
	// +optional
	GPU *GPUSpec `json:"gpu,omitempty"`
	// Image is the container image to use. Overrides Spec.Image if set.
	// +optional
	Image *string `json:"image,omitempty"`
	// ConfigMaps carries information of other ConfigMaps to add to the pod.
	// +optional
	ConfigMaps []NamePath `json:"configMaps,omitempty"`
	// Secrets carries information of secrets to add to the pod.
	// +optional
	Secrets []SecretInfo `json:"secrets,omitempty"`
	// Env carries the environment variables to add to the pod.
	// +optional
	Env []apiv1.EnvVar `json:"env,omitempty"`
	// EnvVars carries the environment variables to add to the pod.
	// Deprecated.
	// +optional
	EnvVars map[string]string `json:"envVars,omitempty"`
	// EnvFrom is a list of sources to populate environment variables in the container.
	// +optional
	EnvFrom []apiv1.EnvFromSource `json:"envFrom,omitempty"`
	// EnvSecretKeyRefs holds a mapping from environment variable names to SecretKeyRefs.
	// Deprecated.
	// +optional
	EnvSecretKeyRefs map[string]NameKey `json:"envSecretKeyRefs,omitempty"`
	// Labels are the Kubernetes labels to be added to the pod.
	// +optional
	Labels map[string]string `json:"labels,omitempty"`
	// Annotations are the Kubernetes annotations to be added to the pod.
	// +optional
	Annotations map[string]string `json:"annotations,omitempty"`
	// VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the main container's filesystem.
	// +optional
	VolumeMounts []apiv1.VolumeMount `json:"volumeMounts,omitempty"`
	// Affinity specifies the affinity/anti-affinity settings for the pod.
	// +optional
	Affinity *apiv1.Affinity `json:"affinity,omitempty"`
	// Tolerations specifies the tolerations listed in ".spec.tolerations" to be applied to the pod.
	// +optional
	Tolerations []apiv1.Toleration `json:"tolerations,omitempty"`
	// SecurityContenxt specifies the PodSecurityContext to apply.
	// +optional
	SecurityContenxt *apiv1.PodSecurityContext `json:"securityContext,omitempty"`
	// SchedulerName specifies the scheduler that will be used for scheduling
	// +optional
	SchedulerName *string `json:"schedulerName,omitempty"`
	// Sidecars is a list of sidecar containers that run along side the main Spark container.
	// +optional
	Sidecars []apiv1.Container `json:"sidecars,omitempty"`
	// InitContainers is a list of init-containers that run to completion before the main Spark container.
	// +optional
	InitContainers []apiv1.Container `json:"initContainers,omitempty"`
	// HostNetwork indicates whether to request host networking for the pod or not.
	// +optional
	HostNetwork *bool `json:"hostNetwork,omitempty"`
	// NodeSelector is the Kubernetes node selector to be added to the driver and executor pods.
	// This field is mutually exclusive with nodeSelector at SparkApplication level (which will be deprecated).
	// +optional
	NodeSelector map[string]string `json:"nodeSelector,omitempty"`
	// DnsConfig dns settings for the pod, following the Kubernetes specifications.
	// +optional
	DNSConfig *apiv1.PodDNSConfig `json:"dnsConfig,omitempty"`
	// NodeName is kubernetes node name to be added to the driver and executor pods.
	// +optional
	NodeName *string `json:"nodeName,omitempty"`
	// DnsPolicy settings for pod.
	// +optional
	DNSPolicy apiv1.DNSPolicy `json:"dnsPolicy,omitempty"`
	// Termination grace periond seconds for the pod
	// +optional
	TerminationGracePeriodSeconds *int64 `json:"terminationGracePeriodSeconds,omitempty"`
	// Resources to override
	// +optional
	CustomResources apiv1.ResourceRequirements `json:"customResources,omitempty"`
}

SparkPodSpec defines common things that can be customized for a Spark driver or executor pod. TODO: investigate if we should use v1.PodSpec and limit what can be set instead.

func (*SparkPodSpec) DeepCopy

func (in *SparkPodSpec) DeepCopy() *SparkPodSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SparkPodSpec.

func (*SparkPodSpec) DeepCopyInto

func (in *SparkPodSpec) DeepCopyInto(out *SparkPodSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL