Package org.apache.flink.runtime.executiongraph

Class IntermediateResult

java.lang.Object

org.apache.flink.runtime.executiongraph.IntermediateResult

public class IntermediateResult
extends Object

Constructor Summary

Constructors

Constructor	Description
IntermediateResult(IntermediateDataSet intermediateDataSet, ExecutionJobVertex producer, int numParallelProducers, ResultPartitionType resultType, ExecutionPlanSchedulingContext executionPlanSchedulingContext)

Method Summary

Modifier and Type	Method	Description
boolean	areAllConsumerVerticesCreated()
CachedShuffleDescriptors	cacheShuffleDescriptors(ConsumedPartitionGroup consumedPartitionGroup, TaskDeploymentDescriptorFactory.ShuffleDescriptorAndIndex[] shuffleDescriptors)
void	clearCachedInformationForPartitionGroup(ConsumedPartitionGroup consumedPartitionGroup)
CachedShuffleDescriptors	getCachedShuffleDescriptors(ConsumedPartitionGroup consumedPartitionGroup)
int	getConnectionIndex()
List<JobVertexID>	getConsumerVertices()
DistributionPattern	getConsumingDistributionPattern()
IntermediateDataSetID	getId()
int	getNumberOfAssignedPartitions()
IntermediateResultPartition	getPartitionById(IntermediateResultPartitionID resultPartitionId)	Returns the partition with the given ID.
IntermediateResultPartition[]	getPartitions()
ExecutionJobVertex	getProducer()
ResultPartitionType	getResultType()
boolean	isBroadcast()	Determines whether the associated intermediate data set uses a broadcast distribution pattern.
boolean	isForward()
boolean	isSingleSubpartitionContainsAllData()	Checks if a single subpartition contains all the produced data.
void	markPartitionFinished(ConsumedPartitionGroup consumedPartitionGroup, IntermediateResultPartition resultPartition)
void	setPartition(int partitionNumber, IntermediateResultPartition partition)
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

IntermediateResult

public IntermediateResult(IntermediateDataSet intermediateDataSet,
                          ExecutionJobVertex producer,
                          int numParallelProducers,
                          ResultPartitionType resultType,
                          ExecutionPlanSchedulingContext executionPlanSchedulingContext)

Method Detail

areAllConsumerVerticesCreated

public boolean areAllConsumerVerticesCreated()

setPartition

public void setPartition(int partitionNumber,
                         IntermediateResultPartition partition)

getId

public IntermediateDataSetID getId()

getProducer

public ExecutionJobVertex getProducer()

getPartitions

public IntermediateResultPartition[] getPartitions()

getConsumerVertices

public List<JobVertexID> getConsumerVertices()

getPartitionById

public IntermediateResultPartition getPartitionById(IntermediateResultPartitionID resultPartitionId)

Returns the partition with the given ID.

Parameters:

resultPartitionId - ID of the partition to look up

Returns:

Intermediate result partition with the given ID

Throws:

NullPointerException - If partition ID null

IllegalArgumentException - Thrown if unknown partition ID

getNumberOfAssignedPartitions

public int getNumberOfAssignedPartitions()

getResultType

public ResultPartitionType getResultType()

getConsumingDistributionPattern

public DistributionPattern getConsumingDistributionPattern()

isBroadcast

public boolean isBroadcast()

Determines whether the associated intermediate data set uses a broadcast distribution pattern.

A broadcast distribution pattern indicates that all data produced by this intermediate data set should be broadcast to every downstream consumer.

Returns:

true if the intermediate data set is using a broadcast distribution pattern; false otherwise.

isForward

public boolean isForward()

isSingleSubpartitionContainsAllData

public boolean isSingleSubpartitionContainsAllData()

Checks if a single subpartition contains all the produced data. This condition indicate that the data was intended to be broadcast to all consumers. If the decision to broadcast was made before the data production, this flag would likely be set accordingly. Conversely, if the broadcasting decision was made post-production, this flag will be false.

Returns:

true if a single subpartition contains all the data; false otherwise.

getConnectionIndex

public int getConnectionIndex()

getCachedShuffleDescriptors

public CachedShuffleDescriptors getCachedShuffleDescriptors(ConsumedPartitionGroup consumedPartitionGroup)

cacheShuffleDescriptors

public CachedShuffleDescriptors cacheShuffleDescriptors(ConsumedPartitionGroup consumedPartitionGroup,
                                                        TaskDeploymentDescriptorFactory.ShuffleDescriptorAndIndex[] shuffleDescriptors)

markPartitionFinished

public void markPartitionFinished(ConsumedPartitionGroup consumedPartitionGroup,
                                  IntermediateResultPartition resultPartition)

clearCachedInformationForPartitionGroup

public void clearCachedInformationForPartitionGroup(ConsumedPartitionGroup consumedPartitionGroup)

toString

public String toString()

Overrides:

toString in class Object