Documentation ¶
Index ¶
- type Scheduler
- func (s *Scheduler) CreateTrainingJob(jobName string)
- func (s *Scheduler) DeleteTrainingJob(jobName string)
- func (s *Scheduler) GetAllTrainingJob() string
- func (s *Scheduler) Run()
- func (s *Scheduler) Stop()
- func (s *Scheduler) TriggerResched()
- func (s *Scheduler) TriggerReschedAtTime(time time.Time)
- type SchedulerMetrics
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Scheduler ¶
type Scheduler struct { SchedulerID string TotalGpus int // Waiting & running jobs ReadyJobsMap map[string]*trainingjob.TrainingJob // Completed & failed jobs DoneJobsMap map[string]*trainingjob.TrainingJob // Number of allocated GPUs of each training job JobNumGPU types.JobScheduleResult // SchedulerLock is used to protect ReadyJobsMap, DoneJobsMap and JobNumGPU SchedulerLock sync.RWMutex // Algorithm to schedule training jobs Algorithm string // SchedulerMetrics contains run-time metrics of the scheduler Metrics SchedulerMetrics Router *mux.Router PlacementManager *placement.PlacementManager // contains filtered or unexported fields }
func NewScheduler ¶
func NewScheduler(id string, kConfig *rest.Config, resume bool, algorithm string, placementEnabled bool, useConfigMapOpt bool) (*Scheduler, error)
NewScheduler creates a new scheduler
func (*Scheduler) CreateTrainingJob ¶
func (*Scheduler) DeleteTrainingJob ¶
func (*Scheduler) GetAllTrainingJob ¶
GetAllTrainingJob lists all training jobs and their scheduler, status, and waiting/running/total time
func (*Scheduler) TriggerResched ¶
func (s *Scheduler) TriggerResched()
func (*Scheduler) TriggerReschedAtTime ¶
type SchedulerMetrics ¶
type SchedulerMetrics struct {
// contains filtered or unexported fields
}
Click to show internal directories.
Click to hide internal directories.