org.apache.ignite.internal.processors.hadoop.jobtracker

Class HadoopJobTracker

java.lang.Object

org.apache.ignite.internal.processors.hadoop.HadoopComponent

org.apache.ignite.internal.processors.hadoop.jobtracker.HadoopJobTracker

public class HadoopJobTracker
extends HadoopComponent

Hadoop job tracker.

Field Summary

Fields inherited from class org.apache.ignite.internal.processors.hadoop.HadoopComponent

ctx, log

Constructor Summary

Constructors

Constructor and Description
HadoopJobTracker()

Method Summary

Modifier and Type	Method and Description
IgniteInternalFuture<?>	finishFuture(HadoopJobId jobId) Gets job finish future.
HadoopJobEx	job(HadoopJobId jobId, HadoopJobInfo jobInfo) Gets or creates job instance.
HadoopCounters	jobCounters(HadoopJobId jobId) Returns job counters.
boolean	killJob(HadoopJobId jobId) Kills job.
void	onExternalMappersInitialized(HadoopJobId jobId, Collection<Integer> reducers, HadoopProcessDescriptor desc) Callback from task executor called when process is ready to received shuffle messages.
void	onKernalStart() Callback invoked when all grid components are started.
void	onKernalStop(boolean cancel) Callback invoked before all grid components are stopped.
void	onTaskFinished(HadoopTaskInfo info, HadoopTaskStatus status) Callback from task executor invoked when a task has been finished.
HadoopMapReducePlan	plan(HadoopJobId jobId) Gets job plan by job ID.
void	start(HadoopContext ctx)
HadoopJobStatus	status(HadoopJobId jobId) Gets hadoop job status for given job ID.
static HadoopJobStatus	status(HadoopJobMetadata meta) Convert Hadoop job metadata to job status.
IgniteInternalFuture<HadoopJobId>	submit(HadoopJobId jobId, HadoopJobInfo info) Submits execution of Hadoop job to grid.

Methods inherited from class org.apache.ignite.internal.processors.hadoop.HadoopComponent

stop

Methods inherited from class java.lang.Object

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

Constructor Detail

HadoopJobTracker

public HadoopJobTracker()

Method Detail

start

public void start(HadoopContext ctx)
           throws IgniteCheckedException

Overrides:

start in class HadoopComponent

Parameters:

ctx - Hadoop context.

Throws:

IgniteCheckedException

onKernalStart

public void onKernalStart()
                   throws IgniteCheckedException

Callback invoked when all grid components are started.

Overrides:

onKernalStart in class HadoopComponent

Throws:

IgniteCheckedException

onKernalStop

public void onKernalStop(boolean cancel)

Callback invoked before all grid components are stopped.

Overrides:

onKernalStop in class HadoopComponent

submit

public IgniteInternalFuture<HadoopJobId> submit(HadoopJobId jobId,
                                                HadoopJobInfo info)

Submits execution of Hadoop job to grid.

Parameters:

jobId - Job ID.

info - Job info.

Returns:

Job completion future.

status

public static HadoopJobStatus status(HadoopJobMetadata meta)

Convert Hadoop job metadata to job status.

Parameters:

meta - Metadata.

Returns:

Status.

status

@Nullable
public HadoopJobStatus status(HadoopJobId jobId)
                                 throws IgniteCheckedException

Gets hadoop job status for given job ID.

Parameters:

jobId - Job ID to get status for.

Returns:

Job status for given job ID or null if job was not found.

Throws:

IgniteCheckedException

finishFuture

@Nullable
public IgniteInternalFuture<?> finishFuture(HadoopJobId jobId)
                                               throws IgniteCheckedException

Gets job finish future.

Parameters:

jobId - Job ID.

Returns:

Finish future or null.

Throws:

IgniteCheckedException - If failed.

plan

public HadoopMapReducePlan plan(HadoopJobId jobId)
                         throws IgniteCheckedException

Gets job plan by job ID.

Parameters:

jobId - Job ID.

Returns:

Job plan.

Throws:

IgniteCheckedException - If failed.

onTaskFinished

public void onTaskFinished(HadoopTaskInfo info,
                           HadoopTaskStatus status)

Callback from task executor invoked when a task has been finished.

Parameters:

info - Task info.

status - Task status.

onExternalMappersInitialized

public void onExternalMappersInitialized(HadoopJobId jobId,
                                         Collection<Integer> reducers,
                                         HadoopProcessDescriptor desc)

Callback from task executor called when process is ready to received shuffle messages.

Parameters:

jobId - Job ID.

reducers - Reducers.

desc - Process descriptor.

job

@Nullable
public HadoopJobEx job(HadoopJobId jobId,
                                 @Nullable
                                 HadoopJobInfo jobInfo)
                          throws IgniteCheckedException

Gets or creates job instance.

Parameters:

jobId - Job ID.

jobInfo - Job info.

Returns:

Job.

Throws:

IgniteCheckedException - If failed.

killJob

public boolean killJob(HadoopJobId jobId)
                throws IgniteCheckedException

Kills job.

Parameters:

jobId - Job ID.

Returns:

True if job was killed.

Throws:

IgniteCheckedException - If failed.

jobCounters

@Nullable
public HadoopCounters jobCounters(HadoopJobId jobId)
                                     throws IgniteCheckedException

Returns job counters.

Parameters:

jobId - Job identifier.

Returns:

Job counters or null if job cannot be found.

Throws:

IgniteCheckedException - If failed.