org.gridbus.broker.scheduler
Class Scheduler

java.lang.Object
  |
  +--org.gridbus.broker.scheduler.Scheduler

All Implemented Interfaces:

java.lang.Runnable

Direct Known Subclasses:

DataScheduler, DBScheduler

public abstract class Scheduler

extends java.lang.Object

implements java.lang.Runnable

This is the base scheduler class, and contains main methods for performing the scheduling/

Version:

2.0

Author:

Srikumar Venugopal (srikumar@cs.mu.oz.au)

Constructor Summary

Scheduler(GridbusFarmingEngine fe)
Constructor for the Scheduler

Method Summary

java.util.Collection	getJobs()
java.util.Collection	getServers() Get the latest servers collection everytime from the farming engine.
java.lang.String	getStats()
boolean	isClassValid() Returns if the current class is a valid one.
boolean	isFailed()
boolean	isFinished() Returns whether the scheduling is finished.
void	recoverStats() This method re-initialises the scheduler stats, when the broker is recovering from a crash.
void	run() The run method enables the scheduler to run as a seperate thread, and calls the 'schedule()' method.
abstract void	schedule() Abstract method: schedule to be implemented by derived classes which implement various scheduling algorithms.
void	setFailed(boolean failed)
void	setFinished(boolean b) Sets the 'finished' state of the scheduler.
void	sortByCost(java.util.Vector server_vect) Sorts the compute server list by cost

Methods inherited from class java.lang.Object

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

Constructor Detail

Scheduler

public Scheduler(GridbusFarmingEngine fe)

Constructor for the Scheduler

Parameters:

fe - - Farming Engine

Method Detail

sortByCost

public void sortByCost(java.util.Vector server_vect)

Sorts the compute server list by cost

Parameters:

server_vect -

recoverStats

public final void recoverStats()

This method re-initialises the scheduler stats, when the broker is recovering from a crash.

getStats

public final java.lang.String getStats()

Returns:

Stats formatted as a string

run

public final void run()

The run method enables the scheduler to run as a seperate thread, and calls the 'schedule()' method.

Specified by:

run in interface java.lang.Runnable

See Also:

Runnable.run()

isFailed

public boolean isFailed()

Returns:

Returns the failed.

setFailed

public void setFailed(boolean failed)

Parameters:

failed - The failed to set.

isClassValid

public boolean isClassValid()

Returns if the current class is a valid one. This is used to avoid class loader problems with Tomcat.

Returns:

boolean

schedule

public abstract void schedule()

Abstract method: schedule to be implemented by derived classes which implement various scheduling algorithms.

isFinished

public boolean isFinished()

Returns whether the scheduling is finished.

Returns:

boolean

setFinished

public void setFinished(boolean b)

Sets the 'finished' state of the scheduler.

Parameters:

b - - boolean

getServers

public java.util.Collection getServers()

Get the latest servers collection everytime from the farming engine. (So that servers can be added / removed from the farming engine dynamically).

Returns:

getJobs

public java.util.Collection getJobs()

Returns: