Package org.apache.hadoop.mapreduce.v2.app.rm.preemption

Class KillAMPreemptionPolicy

java.lang.Object

org.apache.hadoop.mapreduce.v2.app.rm.preemption.KillAMPreemptionPolicy

All Implemented Interfaces:

AMPreemptionPolicy

public class KillAMPreemptionPolicy
extends Object
implements AMPreemptionPolicy

Sample policy that aggressively kills tasks when requested.

Nested Class Summary

Nested classes/interfaces inherited from interface org.apache.hadoop.mapreduce.v2.app.rm.preemption.AMPreemptionPolicy

AMPreemptionPolicy.Context

Constructor Summary

Constructors

Constructor	Description
KillAMPreemptionPolicy()

Method Summary

Modifier and Type	Method	Description
org.apache.hadoop.mapreduce.checkpoint.TaskCheckpointID	getCheckpointID(org.apache.hadoop.mapreduce.v2.api.records.TaskId taskId)	Method to retrieve the latest checkpoint for a given TaskId
void	handleCompletedContainer(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId attemptID)	Callback informing the policy of containers exiting cleanly.
void	handleFailedContainer(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId attemptID)	Callback informing the policy of containers exiting with a failure.
void	init(AppContext context)
boolean	isPreempted(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId yarnAttemptID)	This method is invoked by components interested to learn whether a certain task is being preempted.
void	preempt(AMPreemptionPolicy.Context ctxt, org.apache.hadoop.yarn.api.records.PreemptionMessage preemptionRequests)	Callback informing the policy of ResourceManager. requests for resources to return to the cluster.
void	reportSuccessfulPreemption(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId taskAttemptID)	This method is used to report to the policy that a certain task has been successfully preempted (for bookeeping, counters, etc..)
void	setCheckpointID(org.apache.hadoop.mapreduce.v2.api.records.TaskId taskId, org.apache.hadoop.mapreduce.checkpoint.TaskCheckpointID cid)	Method to store the latest CheckpointID for a given TaskId.

Methods inherited from class java.lang.Object

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

Constructor Details

KillAMPreemptionPolicy

public KillAMPreemptionPolicy()

Method Details

init

public void init(AppContext context)

Specified by:

init in interface AMPreemptionPolicy

preempt

public void preempt(AMPreemptionPolicy.Context ctxt,
 org.apache.hadoop.yarn.api.records.PreemptionMessage preemptionRequests)

Description copied from interface: AMPreemptionPolicy

Callback informing the policy of ResourceManager. requests for resources to return to the cluster. The policy may take arbitrary action to satisfy requests by checkpointing task state, returning containers, or ignoring requests. The RM may elect to enforce these requests by forcibly killing containers not returned after some duration.

Specified by:

preempt in interface AMPreemptionPolicy

Parameters:

ctxt - Handle to the current state of running containers

preemptionRequests - Request from RM for resources to return.

handleFailedContainer

public void handleFailedContainer(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId attemptID)

Description copied from interface: AMPreemptionPolicy

Callback informing the policy of containers exiting with a failure. This allows the policy to implemnt cleanup/compensating actions.

Specified by:

handleFailedContainer in interface AMPreemptionPolicy

Parameters:

attemptID - Task attempt that failed

isPreempted

public boolean isPreempted(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId yarnAttemptID)

Description copied from interface: AMPreemptionPolicy

This method is invoked by components interested to learn whether a certain task is being preempted.

Specified by:

isPreempted in interface AMPreemptionPolicy

Parameters:

yarnAttemptID - Task attempt to query

Returns:

true if this attempt is being preempted

reportSuccessfulPreemption

public void reportSuccessfulPreemption(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId taskAttemptID)

Description copied from interface: AMPreemptionPolicy

This method is used to report to the policy that a certain task has been successfully preempted (for bookeeping, counters, etc..)

Specified by:

reportSuccessfulPreemption in interface AMPreemptionPolicy

Parameters:

taskAttemptID - Task attempt that preempted

getCheckpointID

public org.apache.hadoop.mapreduce.checkpoint.TaskCheckpointID getCheckpointID(org.apache.hadoop.mapreduce.v2.api.records.TaskId taskId)

Description copied from interface: AMPreemptionPolicy

Method to retrieve the latest checkpoint for a given TaskId

Specified by:

getCheckpointID in interface AMPreemptionPolicy

Parameters:

taskId - TaskID

Returns:

CheckpointID associated with this task or null

setCheckpointID

public void setCheckpointID(org.apache.hadoop.mapreduce.v2.api.records.TaskId taskId,
 org.apache.hadoop.mapreduce.checkpoint.TaskCheckpointID cid)

Description copied from interface: AMPreemptionPolicy

Method to store the latest CheckpointID for a given TaskId. Assigning a null is akin to remove all previous checkpoints for this task.

Specified by:

setCheckpointID in interface AMPreemptionPolicy

Parameters:

taskId - TaskID

cid - Checkpoint to assign or null to remove it.

handleCompletedContainer

public void handleCompletedContainer(org.apache.hadoop.mapreduce.v2.api.records.TaskAttemptId attemptID)

Description copied from interface: AMPreemptionPolicy

Callback informing the policy of containers exiting cleanly. This is reported to the policy for bookeeping purposes.

Specified by:

handleCompletedContainer in interface AMPreemptionPolicy

Parameters:

attemptID - Task attempt that completed