Package org.apache.flink.runtime.io.network.partition

Class TaskExecutorPartitionTrackerImpl

java.lang.Object

org.apache.flink.runtime.io.network.partition.AbstractPartitionTracker<org.apache.flink.api.common.JobID,TaskExecutorPartitionInfo>

org.apache.flink.runtime.io.network.partition.TaskExecutorPartitionTrackerImpl

All Implemented Interfaces:

PartitionTracker<org.apache.flink.api.common.JobID,TaskExecutorPartitionInfo>, TaskExecutorPartitionTracker

public class TaskExecutorPartitionTrackerImpl
extends AbstractPartitionTracker<org.apache.flink.api.common.JobID,TaskExecutorPartitionInfo>
implements TaskExecutorPartitionTracker

Utility for tracking partitions and issuing release calls to task executors and shuffle masters.

Field Summary

Fields inherited from class org.apache.flink.runtime.io.network.partition.AbstractPartitionTracker

partitionInfos, partitionTable

Constructor Summary

Constructors

Constructor	Description
TaskExecutorPartitionTrackerImpl(ShuffleEnvironment<?,?> shuffleEnvironment)

Method Summary

Modifier and Type	Method	Description
ClusterPartitionReport	createClusterPartitionReport()	Creates a ClusterPartitionReport, describing which cluster partitions are currently available.
Collection<TaskExecutorPartitionInfo>	getTrackedPartitionsFor(org.apache.flink.api.common.JobID producingJobId)	Get all partitions tracked for the given job.
void	promoteJobPartitions(Collection<ResultPartitionID> partitionsToPromote)	Promotes the given partitions.
void	startTrackingPartition(org.apache.flink.api.common.JobID producingJobId, TaskExecutorPartitionInfo partitionInfo)	Starts the tracking of the given partition for the given job.
void	stopTrackingAndReleaseAllClusterPartitions()	Releases and stops tracking all partitions.
void	stopTrackingAndReleaseClusterPartitions(Collection<IntermediateDataSetID> dataSetsToRelease)	Releases partitions associated with the given datasets and stops tracking of partitions that were released.
void	stopTrackingAndReleaseJobPartitions(Collection<ResultPartitionID> partitionsToRelease)	Releases the given partitions and stop the tracking of partitions that were released.
void	stopTrackingAndReleaseJobPartitionsFor(org.apache.flink.api.common.JobID producingJobId)	Releases all partitions for the given job and stop the tracking of partitions that were released.

Methods inherited from class org.apache.flink.runtime.io.network.partition.AbstractPartitionTracker

isPartitionTracked, isTrackingPartitionsFor, stopTrackingPartitions, stopTrackingPartitionsFor

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.flink.runtime.io.network.partition.PartitionTracker

isPartitionTracked, isTrackingPartitionsFor, stopTrackingPartitions, stopTrackingPartitionsFor

Constructor Details

TaskExecutorPartitionTrackerImpl

public TaskExecutorPartitionTrackerImpl(ShuffleEnvironment<?,?> shuffleEnvironment)

Method Details

startTrackingPartition

public void startTrackingPartition(org.apache.flink.api.common.JobID producingJobId,
 TaskExecutorPartitionInfo partitionInfo)

Description copied from interface: TaskExecutorPartitionTracker

Starts the tracking of the given partition for the given job.

Specified by:

startTrackingPartition in interface TaskExecutorPartitionTracker

Parameters:

producingJobId - ID of job by which the partition is produced

partitionInfo - information about the partition

stopTrackingAndReleaseJobPartitions

public void stopTrackingAndReleaseJobPartitions(Collection<ResultPartitionID> partitionsToRelease)

Description copied from interface: TaskExecutorPartitionTracker

Releases the given partitions and stop the tracking of partitions that were released.

Specified by:

stopTrackingAndReleaseJobPartitions in interface TaskExecutorPartitionTracker

stopTrackingAndReleaseJobPartitionsFor

public void stopTrackingAndReleaseJobPartitionsFor(org.apache.flink.api.common.JobID producingJobId)

Description copied from interface: TaskExecutorPartitionTracker

Releases all partitions for the given job and stop the tracking of partitions that were released.

Specified by:

stopTrackingAndReleaseJobPartitionsFor in interface TaskExecutorPartitionTracker

getTrackedPartitionsFor

public Collection<TaskExecutorPartitionInfo> getTrackedPartitionsFor(org.apache.flink.api.common.JobID producingJobId)

Description copied from interface: TaskExecutorPartitionTracker

Get all partitions tracked for the given job.

Specified by:

getTrackedPartitionsFor in interface TaskExecutorPartitionTracker

Parameters:

producingJobId - the job id

Returns:

the tracked partitions

promoteJobPartitions

public void promoteJobPartitions(Collection<ResultPartitionID> partitionsToPromote)

Description copied from interface: TaskExecutorPartitionTracker

Promotes the given partitions.

Specified by:

promoteJobPartitions in interface TaskExecutorPartitionTracker

stopTrackingAndReleaseClusterPartitions

public void stopTrackingAndReleaseClusterPartitions(Collection<IntermediateDataSetID> dataSetsToRelease)

Description copied from interface: TaskExecutorPartitionTracker

Releases partitions associated with the given datasets and stops tracking of partitions that were released.

Specified by:

stopTrackingAndReleaseClusterPartitions in interface TaskExecutorPartitionTracker

Parameters:

dataSetsToRelease - data sets to release

stopTrackingAndReleaseAllClusterPartitions

public void stopTrackingAndReleaseAllClusterPartitions()

Description copied from interface: TaskExecutorPartitionTracker

Releases and stops tracking all partitions.

Specified by:

stopTrackingAndReleaseAllClusterPartitions in interface TaskExecutorPartitionTracker

createClusterPartitionReport

public ClusterPartitionReport createClusterPartitionReport()

Description copied from interface: TaskExecutorPartitionTracker

Creates a ClusterPartitionReport, describing which cluster partitions are currently available.

Specified by:

createClusterPartitionReport in interface TaskExecutorPartitionTracker