edu.rice.hj.runtime.mapreduce2

Class MapReduceTask<K1,V1,K2,V2,V3>

java.lang.Object

edu.rice.hj.runtime.mapreduce2.MapReduceTask<K1,V1,K2,V2,V3>

Direct Known Subclasses:

WordCount

public abstract class MapReduceTask<K1,V1,K2,V2,V3>
extends Object

map(k1, v1) -> list(k2, v2) reduce(k2, list(v2)) -> list(v3)

Author:

Shams Imam (shams@rice.edu)

Field Summary

Fields

Modifier and Type	Field and Description
protected int	numMapTasks
protected int	numReduceTasks
protected boolean	printTimes

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	MapReduceTask(int numMapTasks, int numReduceTasks, boolean printTimes) Constructor for MapReduceTask.

Method Summary

Modifier and Type	Method and Description
abstract void	map(K1 inKey, V1 inValue, Aggregator<K2,V2> aggregator) The MAP function.
abstract Map<K1,List<V1>>	partition(int partitionIndex) PARTITION the input to be divided into the MAP tasks.
abstract V3	reduce(K2 inKey, List<V2> inValue, V3 initialValue) The REDUCE function.
abstract V3	reduceIdentity(K2 inKey) reduceIdentity.
List<V3>	results() results.
void	run() run.

Methods inherited from class java.lang.Object

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

Field Detail

numMapTasks

protected final int numMapTasks

numReduceTasks

protected final int numReduceTasks

printTimes

protected final boolean printTimes

Constructor Detail

MapReduceTask

protected MapReduceTask(int numMapTasks,
                        int numReduceTasks,
                        boolean printTimes)

Constructor for MapReduceTask.

Parameters:

numMapTasks - a int.

numReduceTasks - a int.

printTimes - a boolean.

Method Detail

results

public final List<V3> results()

results.

Returns:

a List object.

run

public final void run()
               throws SuspendableException

run.

Throws:

SuspendableException

partition

public abstract Map<K1,List<V1>> partition(int partitionIndex)

PARTITION the input to be divided into the MAP tasks. This should be a constant time operation to avoid serialization bottlenecks.

Parameters:

partitionIndex - The index of partition to create.

Returns:

array representing each of the partitions.

map

public abstract void map(K1 inKey,
                         V1 inValue,
                         Aggregator<K2,V2> aggregator)

The MAP function.

Parameters:

inKey - The input key.

inValue - The input value.

aggregator - Stores the intermediates.

reduceIdentity

public abstract V3 reduceIdentity(K2 inKey)

reduceIdentity.

Parameters:

inKey - The input key.

Returns:

the identity value for the reducer.

reduce

public abstract V3 reduce(K2 inKey,
                          List<V2> inValue,
                          V3 initialValue)

The REDUCE function.

Parameters:

inKey - The input key.

inValue - The input value.

initialValue - The current value representing the given key.

Returns:

the reduced value.