teetime.stage.taskfarm.adaptation.analysis

Class AbstractThroughputAlgorithm

java.lang.Object

teetime.stage.taskfarm.adaptation.analysis.AbstractThroughputAlgorithm

Direct Known Subclasses:

MeanAlgorithm, RegressionAlgorithm, WeightedAlgorithm

public abstract class AbstractThroughputAlgorithm
extends Object

Represents the calculation of the throughput score by means of the prediction value calculated by the concrete Throughput Algorithm.

Author:

Christian Claus Wiechmann

Field Summary

Fields

Modifier and Type	Field and Description
static int	INVALID_SCORE represents an invalid throughput score
protected int	window Number of recently measured throughput values that we will analyze.

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	AbstractThroughputAlgorithm(TaskFarmConfiguration<?,?,?> configuration) Constructor.

Method Summary

Modifier and Type	Method and Description
protected abstract double	doAnalysis(ThroughputHistory history) Calculates a prediction for the most recent ThroughputEntry by means of the last "window" entries.
double	getTroughputAnalysis(ThroughputHistory history) Calculates the throughput score for a given history.

Methods inherited from class java.lang.Object

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

Field Detail

INVALID_SCORE

public static final int INVALID_SCORE

represents an invalid throughput score

See Also:

Constant Field Values

window

protected final int window

Number of recently measured throughput values that we will analyze.

Constructor Detail

AbstractThroughputAlgorithm

protected AbstractThroughputAlgorithm(TaskFarmConfiguration<?,?,?> configuration)

Constructor.

Parameters:

configuration - TaskFarmConfiguration of the Task Farm which this algorithm is used for

Method Detail

doAnalysis

protected abstract double doAnalysis(ThroughputHistory history)

Calculates a prediction for the most recent ThroughputEntry by means of the last "window" entries.

Parameters:

history - measured throughput values + timestamps

Returns:

throughput prediction for the most recent value

getTroughputAnalysis

public final double getTroughputAnalysis(ThroughputHistory history)

Calculates the throughput score for a given history.

Parameters:

history - history of measured values

Returns:

throughput score