be.ac.ua.broker.scheduling.workload.wm1
Class SuperSpotTHOptimizer

java.lang.Object
  be.ac.ua.broker.scheduling.workload.wm1.SuperSpotTHOptimizer

All Implemented Interfaces:

Optimizer<WorkloadWM1>

public class SuperSpotTHOptimizer
extends java.lang.Object
implements Optimizer<WorkloadWM1>

This class is used to perform the spot-optimized scheduling for workload model one tasks. Spot-Optimized scheduling is an alteration of the optimized algorithm, that ensures that the least loaded resources contain as much spot-enabled task parts as possible.

Author:

Vermeersch Kurt

Constructor Summary

SuperSpotTHOptimizer()

Method Summary

void	add(WorkloadWM1 wl) Add the provided workload to the scheduling optimizer.
java.util.List<ScheduledTaskCollection>	determineOptimal()
java.util.List<WorkloadWM1>	getWorkLoadsForDeadline(int deadline, boolean isSpotEnabled) Getter for the tasks begin scheduled which have a given deadline.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

SuperSpotTHOptimizer

public SuperSpotTHOptimizer()

Method Detail

add

public void add(WorkloadWM1 wl)

Description copied from interface: Optimizer

Add the provided workload to the scheduling optimizer.

Specified by:

add in interface Optimizer<WorkloadWM1>

Parameters:

wl - the workload to add.

determineOptimal

public java.util.List<ScheduledTaskCollection> determineOptimal()

Specified by:

determineOptimal in interface Optimizer<WorkloadWM1>

Returns:

a list of scheduledTaskCollections, one of these represents the schedule containing subTasks for a single instance. This schedules are made in a way that tries to minimize the total cost of running the workload on EC2.

getWorkLoadsForDeadline

public java.util.List<WorkloadWM1> getWorkLoadsForDeadline(int deadline,
                                                           boolean isSpotEnabled)

Getter for the tasks begin scheduled which have a given deadline.

Parameters:

deadline - the deadline of which we want to get the tasks.

Returns:

the list of workloads associated with the given deadline