v1alpha2

Documentation

Index

• Constants
• type MPIJobController
  • func NewMPIJobController(kubeClient kubernetes.Interface, kubeflowClient clientset.Interface, ...) *MPIJobController
  • func (c *MPIJobController) Run(threadiness int, stopCh <-chan struct{}) error

Constants

const (
	// ErrResourceExists is used as part of the Event 'reason' when an MPIJob
	// fails to sync due to dependent resources of the same name already
	// existing.
	ErrResourceExists = "ErrResourceExists"

	// MessageResourceExists is the message used for Events when a resource
	// fails to sync due to dependent resources already existing.
	MessageResourceExists = "Resource %q of Kind %q already exists and is not managed by MPIJob"

	// ErrResourceDoesNotExist is used as part of the Event 'reason' when some
	// resource is missing in yaml
	ErrResourceDoesNotExist = "ErrResourceDoesNotExist"

	// MessageResourceDoesNotExist is used for Events when some
	// resource is missing in yaml
	MessageResourceDoesNotExist = "Resource %q is missing in yaml"
)

Types

type MPIJobController

type MPIJobController struct {
	// contains filtered or unexported fields
}

MPIJobController is the controller implementation for MPIJob resources.

func NewMPIJobController

func NewMPIJobController(
	kubeClient kubernetes.Interface,
	kubeflowClient clientset.Interface,
	kubeBatchClientSet kubebatchclient.Interface,
	configMapInformer coreinformers.ConfigMapInformer,
	serviceAccountInformer coreinformers.ServiceAccountInformer,
	roleInformer rbacinformers.RoleInformer,
	roleBindingInformer rbacinformers.RoleBindingInformer,
	statefulSetInformer appsinformers.StatefulSetInformer,
	jobInformer batchinformers.JobInformer,
	podgroupsInformer podgroupsinformer.PodGroupInformer,
	mpiJobInformer informers.MPIJobInformer,
	kubectlDeliveryImage string,
	gangSchedulerName string) *MPIJobController

NewMPIJobController returns a new MPIJob controller.

func (*MPIJobController) Run

func (c *MPIJobController) Run(threadiness int, stopCh <-chan struct{}) error

Run will set up the event handlers for types we are interested in, as well as syncing informer caches and starting workers. It will block until stopCh is closed, at which point it will shutdown the work queue and wait for workers to finish processing their current work items.