org.apache.ignite.internal.processors.hadoop

Interface HadoopJobInfo

All Known Implementing Classes:

HadoopDefaultJobInfo

public interface HadoopJobInfo

Extended job description.

Method Summary

Methods

Modifier and Type	Method and Description
HadoopJobEx	createJob(Class<? extends HadoopJobEx> jobCls, HadoopJobId jobId, IgniteLogger log, String[] libNames, HadoopHelper helper) Creates new job instance for the given ID.
boolean	hasCombiner() Checks whether job has combiner.
boolean	hasReducer() Checks whether job has reducer.
String	jobName() Gets job name.
String	property(String name) Gets optional configuration property for the job.
int	reducers()
String	user() Gets user name.

Method Detail

property

@Nullable
String property(String name)

Gets optional configuration property for the job.

Parameters:

name - Property name.

Returns:

Value or null if none.

hasCombiner

boolean hasCombiner()

Checks whether job has combiner.

Returns:

true If job has combiner.

hasReducer

boolean hasReducer()

Checks whether job has reducer. Actual number of reducers will be in HadoopMapReducePlan.reducers().

Returns:

Number of reducer.

reducers

int reducers()

Returns:

Number of reducers configured for job.

jobName

String jobName()

Gets job name.

Returns:

Job name.

user

String user()

Gets user name.

Returns:

User name.

createJob

HadoopJobEx createJob(Class<? extends HadoopJobEx> jobCls,
                    HadoopJobId jobId,
                    IgniteLogger log,
                    @Nullable
                    String[] libNames,
                    HadoopHelper helper)
                      throws IgniteCheckedException

Creates new job instance for the given ID. HadoopJobInfo is reusable for multiple jobs while HadoopJobEx is for one job execution. This method will be called once for the same ID on one node, though it can be called on the same host multiple times from different processes (in case of multiple nodes on the same host or external execution).

Parameters:

jobCls - The job class.

jobId - Job ID.

log - Logger.

libNames - Optional additional native library names.

helper - HadoopHelper.

Returns:

Job.

Throws:

IgniteCheckedException - If failed.