org.apache.nutch.mapReduce
Class JobTracker.JobInProgress

java.lang.Object
  org.apache.nutch.mapReduce.JobTracker.JobInProgress

Enclosing class:

JobTracker

public class JobTracker.JobInProgress

extends Object

Constructor Summary

JobTracker.JobInProgress(String jobFile)
Create a 'JobInProgress' object, which contains both JobProfile and JobStatus.

Method Summary

int	attemptedMaps()
int	attemptedReduces()
int	completedMaps()
float	completedRatio()
int	completedReduces()
void	completedTask(String taskid) A task assigned to this JobInProgress has reported in successfully.
int	desiredMaps()
int	desiredReduces()
void	failedTask(String taskid) A task assigned to this JobInProgress has reported in as failed.
void	garbageCollect() The job is dead.
long	getFinishTime()
JobProfile	getProfile()
long	getStartTime()
JobStatus	getStatus()
Task	getTask(String taskid)
TaskStatus	getTaskStatus(String taskid)
void	kill() Kill the job and all its component tasks.
void	launch() Start up the tasks
MapOutputLocation[]	locateTasks(String[] taskIds) Return locations for all the indicated taskIds.
void	updateTaskStatus(String taskid, TaskStatus status)

Methods inherited from class java.lang.Object

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

Constructor Detail

JobTracker.JobInProgress

public JobTracker.JobInProgress(String jobFile)
                         throws IOException

Create a 'JobInProgress' object, which contains both JobProfile and JobStatus. Those two sub-objects are sometimes shipped outside of the JobTracker. But JobInProgress adds info that's useful for the JobTracker alone.

Method Detail

launch

public void launch()
            throws IOException

Start up the tasks

Throws:

IOException

kill

public void kill()

Kill the job and all its component tasks.

garbageCollect

public void garbageCollect()
                    throws IOException

The job is dead. We're now GC'ing it, getting rid of the job from all tables. Be sure to remove all of this job's tasks from the various tables.

Throws:

IOException

completedTask

public void completedTask(String taskid)

A task assigned to this JobInProgress has reported in successfully. This might prompt us to launch more tasks, or it might even indicate the job is now complete.

failedTask

public void failedTask(String taskid)

A task assigned to this JobInProgress has reported in as failed. Most of the time, we'll just reschedule execution. However, after many repeated failures we may instead decide to allow the entire job to fail. Even if a task has reported as completed in the past, it might later be reported as failed. That's because the TaskTracker that hosts a map task might die before the entire job can complete. If that happens, we need to schedule reexecution so that downstream reduce tasks can obtain the map task's output.

locateTasks

public MapOutputLocation[] locateTasks(String[] taskIds)

Return locations for all the indicated taskIds. If the task is not complete, don't return anything.

getProfile

public JobProfile getProfile()

getStatus

public JobStatus getStatus()

getTask

public Task getTask(String taskid)

completedRatio

public float completedRatio()

getStartTime

public long getStartTime()

getFinishTime

public long getFinishTime()

desiredMaps

public int desiredMaps()

attemptedMaps

public int attemptedMaps()

completedMaps

public int completedMaps()

desiredReduces

public int desiredReduces()

attemptedReduces

public int attemptedReduces()

completedReduces

public int completedReduces()

updateTaskStatus

public void updateTaskStatus(String taskid,
                             TaskStatus status)

getTaskStatus

public TaskStatus getTaskStatus(String taskid)