Package com.hazelcast.mapreduce.impl.operation

This package contains all remote operations that are needed to control work on supervising or worker nodes.

See:
Description

Class Summary

CancelJobSupervisorOperation	This operation is fired by the jobs owner node to remotely cancel the defined jobId on all nodes.
FireNotificationOperation	This operation is used to transport and execute a notification on a remote node since the current event service is not capable of reliable transmission so this is a fallback to sync operations which is hopefully only a temporary workaround!
GetResultOperation	This operation is used to retrieve results from a remote node
GetResultOperationFactory	This class is used to create GetResultOperation instances. In difference to other implementations of OperationFactory this class is never ever serialized and the DataSerializable methods GetResultOperationFactory.readData(com.hazelcast.nio.ObjectDataInput) and GetResultOperationFactory.writeData(com.hazelcast.nio.ObjectDataOutput) throw UnsupportedOperationExceptions.
KeysAssignmentOperation	This operation is used to request assignment for keys on the job owners node.
KeysAssignmentResult	This class is used to store assignment results in KeysAssignmentOperation executions.
KeyValueJobOperation<K,V>	This operation is used to prepare a KeyValueSource based map reduce operation on all cluster members.
NotifyRemoteExceptionOperation	This operation is used to inform the job owner of a remote exception
PostPonePartitionProcessing	This operation is used to tell the job owner to postpone a mapping phase for the defined partitionId.
ProcessingOperation	This is the base class for all map reduce framework operations, it always contains the name of the JobTracker and the unique jobId
ProcessStatsUpdateOperation	This operation is used to update the process statistics on the owner node
RequestMemberIdAssignment	This operation is used to do some kind of partitionId based processing on non partition based implementations of KeyValueSource (not implementing PartitionIdAware) which can happen for custom data sources like distributed filesystems that are up to the end user on how to manage the distribution.
RequestPartitionMapping	This operation requests a new partition to process by the requester on the job owning node
RequestPartitionProcessed	This operation tells the job owner to mark a partition as fully processed
RequestPartitionReducing	This operation tells the job owner to start reducing phase of the defined partition on the first occurrence of a full emitted chunk
RequestPartitionResult	This class is used to store the requested result of all kinds of processing operations. By default it holds a basic result state and maybe the partitionId it was operated otherwise it's value is defined as -1
StartProcessingJobOperation<K>	This operation is used to start the actual processing after all node prepared to execute the map reduce job

Enum Summary

RequestPartitionResult.ResultState

This enum is used to define the basic state of an operations result

Package com.hazelcast.mapreduce.impl.operation Description

This package contains all remote operations that are needed to control work on supervising or worker nodes.