v1beta1

package
v1.1.0 Latest Latest
Warning

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

Go to latest
Published: Mar 25, 2021 License: Apache-2.0 Imports: 5 Imported by: 0

Documentation

Overview

Package v1alpha2 is the v1alpha2 version of the API. +groupName=kubeflow.org

Index

Constants

View Source
const (
	// EnvKubeflowNamespace is ENV for kubeflow namespace specified by user.
	EnvKubeflowNamespace = "KUBEFLOW_NAMESPACE"

	// DefaultPortName is name of the port used to communicate between scheduler and
	// servers & workers.
	DefaultPortName = "mxjob-port"
	// DefaultContainerName is the name of the MXJob container.
	DefaultContainerName = "mxnet"
	// DefaultPort is default value of the port.
	DefaultPort = 9091
	// DefaultRestartPolicy is default RestartPolicy for MXReplicaSpec.
	DefaultRestartPolicy = RestartPolicyNever
)
View Source
const (
	// GroupName is the group name use in this package.
	GroupName = "kubeflow.org"
	// Kind is the kind name.
	Kind = "MXJob"
	// GroupVersion is the version.
	GroupVersion = "v1beta1"
	// Plural is the Plural for MXJob.
	Plural = "mxjobs"
	// Singular is the singular for MXJob.
	Singular = "mxjob"
	// MXCRD is the CRD name for MXJob.
	MXCRD = "mxjobs.kubeflow.org"
)

Variables

View Source
var (
	// TODO: move SchemeBuilder with zz_generated.deepcopy.go to k8s.io/api.
	// localSchemeBuilder and AddToScheme will stay in k8s.io/kubernetes.
	SchemeBuilder runtime.SchemeBuilder

	AddToScheme = localSchemeBuilder.AddToScheme
)
View Source
var (
	// SchemeGroupVersion is the group version used to register these objects.
	SchemeGroupVersion = schema.GroupVersion{Group: GroupName, Version: GroupVersion}
	// SchemeGroupVersionKind is the GroupVersionKind of the resource.
	SchemeGroupVersionKind = SchemeGroupVersion.WithKind(Kind)
)

Functions

func Int32

func Int32(v int32) *int32

Int32 is a helper routine that allocates a new int32 value to store v and returns a pointer to it.

func IsScheduler

func IsScheduler(typ MXReplicaType) bool

IsScheduler returns true if the type is Scheduler.

func RegisterDefaults

func RegisterDefaults(scheme *runtime.Scheme) error

RegisterDefaults adds defaulters functions to the given scheme. Public to allow building arbitrary schemes. All generated defaulters are covering - they call all nested defaulters.

func Resource

func Resource(resource string) schema.GroupResource

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

func SetDefaults_MXJob

func SetDefaults_MXJob(mxjob *MXJob)

SetDefaults_MXJob sets any unspecified values to defaults.

func SetObjectDefaults_MXJob

func SetObjectDefaults_MXJob(in *MXJob)

func SetObjectDefaults_MXJobList

func SetObjectDefaults_MXJobList(in *MXJobList)

Types

type CleanPodPolicy

type CleanPodPolicy string

CleanPodPolicy describes how to deal with pods when the MXJob is finished.

const (
	CleanPodPolicyUndefined CleanPodPolicy = ""
	CleanPodPolicyAll       CleanPodPolicy = "All"
	CleanPodPolicyRunning   CleanPodPolicy = "Running"
	CleanPodPolicyNone      CleanPodPolicy = "None"
)

type JobModeType

type JobModeType string

JobModeType id the type for JobMode

const (
	// Train Mode, in this mode requested MXReplicaSpecs need
	// has Server, Scheduler, Worker
	MXTrain JobModeType = "MXTrain"

	// Tune Mode, in this mode requested MXReplicaSpecs need
	// has Tuner
	MXTune JobModeType = "MXTune"
)

type MXJob

type MXJob struct {
	metav1.TypeMeta `json:",inline"`

	// Standard object's metadata.
	metav1.ObjectMeta `json:"metadata,omitempty"`

	// Specification of the desired behavior of the MXJob.
	Spec MXJobSpec `json:"spec,omitempty"`

	// Most recently observed status of the MXJob.
	// This data may not be up to date.
	// Populated by the system.
	// Read-only.
	Status MXJobStatus `json:"status,omitempty"`
}

MXJob represents the configuration of signal MXJob

func (*MXJob) DeepCopy

func (in *MXJob) DeepCopy() *MXJob

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

func (*MXJob) DeepCopyInto

func (in *MXJob) DeepCopyInto(out *MXJob)

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

func (*MXJob) DeepCopyObject

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

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

type MXJobCondition

type MXJobCondition struct {
	// Type of MXJob condition.
	Type MXJobConditionType `json:"type"`
	// Status of the condition, one of True, False, Unknown.
	Status v1.ConditionStatus `json:"status"`
	// The reason for the condition's last transition.
	Reason string `json:"reason,omitempty"`
	// A human readable message indicating details about the transition.
	Message string `json:"message,omitempty"`
	// The last time this condition was updated.
	LastUpdateTime metav1.Time `json:"lastUpdateTime,omitempty"`
	// Last time the condition transitioned from one status to another.
	LastTransitionTime metav1.Time `json:"lastTransitionTime,omitempty"`
}

MXJobCondition describes the state of the MXJob at a certain point.

func (*MXJobCondition) DeepCopy

func (in *MXJobCondition) DeepCopy() *MXJobCondition

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

func (*MXJobCondition) DeepCopyInto

func (in *MXJobCondition) DeepCopyInto(out *MXJobCondition)

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

type MXJobConditionType

type MXJobConditionType string

MXJobConditionType defines all kinds of types of MXJobStatus.

const (
	// MXJobCreated means the mxjob has been accepted by the system,
	// but one or more of the pods/services has not been started.
	// This includes time before pods being scheduled and launched.
	MXJobCreated MXJobConditionType = "Created"

	// MXJobRunning means all sub-resources (e.g. services/pods) of this MXJob
	// have been successfully scheduled and launched.
	// The training is running without error.
	MXJobRunning MXJobConditionType = "Running"

	// MXJobRestarting means one or more sub-resources (e.g. services/pods) of this MXJob
	// reached phase failed but maybe restarted according to it's restart policy
	// which specified by user in v1.PodTemplateSpec.
	// The training is freezing/pending.
	MXJobRestarting MXJobConditionType = "Restarting"

	// MXJobSucceeded means all sub-resources (e.g. services/pods) of this MXJob
	// reached phase have terminated in success.
	// The training is complete without error.
	MXJobSucceeded MXJobConditionType = "Succeeded"

	// MXJobFailed means one or more sub-resources (e.g. services/pods) of this MXJob
	// reached phase failed with no restarting.
	// The training has failed its execution.
	MXJobFailed MXJobConditionType = "Failed"
)

type MXJobList

type MXJobList struct {
	metav1.TypeMeta `json:",inline"`

	// Standard list metadata.
	metav1.ListMeta `json:"metadata,omitempty"`

	// List of MXJobs.
	Items []MXJob `json:"items"`
}

MXJobList is a list of MXJobs.

func (*MXJobList) DeepCopy

func (in *MXJobList) DeepCopy() *MXJobList

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

func (*MXJobList) DeepCopyInto

func (in *MXJobList) DeepCopyInto(out *MXJobList)

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

func (*MXJobList) DeepCopyObject

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

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

type MXJobSpec

type MXJobSpec struct {
	// CleanPodPolicy defines the policy to kill pods after MXJob is
	// succeeded.
	// Default to All.
	CleanPodPolicy *CleanPodPolicy `json:"cleanPodPolicy,omitempty"`

	// SuccessPolicy defines the policy to mark the MXJob as succeeded.
	// Default to "", using the default rules. When either scheduler
	// or all workers have completed, consider the job is finished.
	SuccessPolicy *SuccessPolicy `json:"successPolicy,omitempty"`

	// TTLSecondsAfterFinished is the TTL to clean up mxnet-jobs (temporary
	// before kubernetes adds the cleanup controller).
	// It may take extra ReconcilePeriod seconds for the cleanup, since
	// reconcile gets called periodically.
	// Default to infinite.
	TTLSecondsAfterFinished *int32 `json:"ttlSecondsAfterFinished,omitempty"`

	// JobMode specify the kind of MXjob to do. Different mode may have
	// different MXReplicaSpecs request
	JobMode JobModeType `json:"jobMode"`

	// MXReplicaSpecs is map of MXReplicaType and MXReplicaSpec
	// specifies the MX replicas to run.
	// For example,
	//   {
	//     "Scheduler": MXReplicaSpec,
	//     "Server": MXReplicaSpec,
	//     "Worker": MXReplicaSpec,
	//   }
	MXReplicaSpecs map[MXReplicaType]*MXReplicaSpec `json:"mxReplicaSpecs"`
}

MXJobSpec is a desired state description of the MXJob.

func (*MXJobSpec) DeepCopy

func (in *MXJobSpec) DeepCopy() *MXJobSpec

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

func (*MXJobSpec) DeepCopyInto

func (in *MXJobSpec) DeepCopyInto(out *MXJobSpec)

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

type MXJobStatus

type MXJobStatus struct {
	// Conditions is an array of current observed MXJob conditions.
	Conditions []MXJobCondition `json:"conditions"`

	// MXReplicaStatuses is map of MXReplicaType and MXReplicaStatus,
	// specifies the status of each MXReplica.
	MXReplicaStatuses map[MXReplicaType]*MXReplicaStatus `json:"mxReplicaStatuses"`

	// Represents time when the MXJob was acknowledged by the MXJob controller.
	// It is not guaranteed to be set in happens-before order across separate operations.
	// It is represented in RFC3339 form and is in UTC.
	StartTime *metav1.Time `json:"startTime,omitempty"`

	// Represents time when the MXJob was completed. It is not guaranteed to
	// be set in happens-before order across separate operations.
	// It is represented in RFC3339 form and is in UTC.
	CompletionTime *metav1.Time `json:"completionTime,omitempty"`

	// Represents last time when the MXJob was reconciled. It is not guaranteed to
	// be set in happens-before order across separate operations.
	// It is represented in RFC3339 form and is in UTC.
	LastReconcileTime *metav1.Time `json:"lastReconcileTime,omitempty"`
}

MXJobStatus represents the current observed state of the MXJob.

func (*MXJobStatus) DeepCopy

func (in *MXJobStatus) DeepCopy() *MXJobStatus

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

func (*MXJobStatus) DeepCopyInto

func (in *MXJobStatus) DeepCopyInto(out *MXJobStatus)

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

type MXReplicaSpec

type MXReplicaSpec struct {
	// Replicas is the desired number of replicas of the given template.
	// If unspecified, defaults to 1.
	Replicas *int32 `json:"replicas,omitempty"`

	// Label is used as tunerServerKey, it's designed for tvm auto-tuning.
	Label string `json:"label,omitempty"`

	// Template is the object that describes the pod that
	// will be created for this MXReplica. RestartPolicy in PodTemplateSpec
	// will be overide by RestartPolicy in MXReplicaSpec
	Template v1.PodTemplateSpec `json:"template,omitempty"`

	// Restart policy for all MXReplicas within the MXJob.
	// One of Always, OnFailure, Never and ExitCode.
	// Default to Never.
	RestartPolicy RestartPolicy `json:"restartPolicy,omitempty"`
}

MXReplicaSpec is a description of the MXReplica

func (*MXReplicaSpec) DeepCopy

func (in *MXReplicaSpec) DeepCopy() *MXReplicaSpec

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

func (*MXReplicaSpec) DeepCopyInto

func (in *MXReplicaSpec) DeepCopyInto(out *MXReplicaSpec)

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

type MXReplicaStatus

type MXReplicaStatus struct {
	// The number of actively running pods.
	Active int32 `json:"active,omitempty"`

	// The number of pods which reached phase Succeeded.
	Succeeded int32 `json:"succeeded,omitempty"`

	// The number of pods which reached phase Failed.
	Failed int32 `json:"failed,omitempty"`
}

MXReplicaStatus represents the current observed state of the MXReplica.

func (*MXReplicaStatus) DeepCopy

func (in *MXReplicaStatus) DeepCopy() *MXReplicaStatus

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

func (*MXReplicaStatus) DeepCopyInto

func (in *MXReplicaStatus) DeepCopyInto(out *MXReplicaStatus)

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

type MXReplicaType

type MXReplicaType string

MXReplicaType is the type for MXReplica.

const (
	// MXReplicaTypeScheduler is the type for scheduler replica in MXNet.
	MXReplicaTypeScheduler MXReplicaType = "Scheduler"

	// MXReplicaTypeServer is the type for parameter servers of distributed MXNet.
	MXReplicaTypeServer MXReplicaType = "Server"

	// MXReplicaTypeWorker is the type for workers of distributed MXNet.
	// This is also used for non-distributed MXNet.
	MXReplicaTypeWorker MXReplicaType = "Worker"

	// MXReplicaTypeTunerTracker
	// This the auto-tuning tracker e.g. autotvm tracker, it will dispatch tuning task to TunerServer
	MXReplicaTypeTunerTracker MXReplicaType = "TunerTracker"

	// MXReplicaTypeTunerServer
	MXReplicaTypeTunerServer MXReplicaType = "TunerServer"

	// MXReplicaTuner is the type for auto-tuning of distributed MXNet.
	// This is also used for non-distributed MXNet.
	MXReplicaTypeTuner MXReplicaType = "Tuner"
)

type RestartPolicy

type RestartPolicy string

RestartPolicy describes how the MXReplicas should be restarted. Only one of the following restart policies may be specified. If none of the following policies is specified, the default one is RestartPolicyAlways.

const (
	RestartPolicyAlways    RestartPolicy = "Always"
	RestartPolicyOnFailure RestartPolicy = "OnFailure"
	RestartPolicyNever     RestartPolicy = "Never"

	// `ExitCode` policy means that user should add exit code by themselves,
	// `mxnet-operator` will check these exit codes to
	// determine the behavior when an error occurs:
	// - 1-127: permanent error, do not restart.
	// - 128-255: retryable error, will restart the pod.
	RestartPolicyExitCode RestartPolicy = "ExitCode"
)

type SuccessPolicy added in v1.1.0

type SuccessPolicy string

SuccessPolicy will determine the success of the job

const (
	SuccessPolicyDefault    SuccessPolicy = ""
	SuccessPolicyAllWorkers SuccessPolicy = "AllWorkers"
)

Jump to

Keyboard shortcuts

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