Module org.elasticsearch.server

Package org.elasticsearch.index.reindex

Class BulkByScrollTask

java.lang.Object

org.elasticsearch.tasks.Task

org.elasticsearch.tasks.CancellableTask

org.elasticsearch.index.reindex.BulkByScrollTask

All Implemented Interfaces:

Traceable

public class BulkByScrollTask
extends CancellableTask

Task storing information about a currently running BulkByScroll request.

When the request is not sliced, this task is the only task created, and starts an action to perform search requests.

When the request is sliced, this task can either represent a coordinating task (using setWorkerCount(int)) or a worker task that performs search queries (using setWorker(float, Integer)).

We don't always know if this task will be a leader or worker task when it's created, because if slices is set to "auto" it may be either depending on the number of shards in the source indices. We figure that out when the request is handled and set it on this class with setWorkerCount(int) or setWorker(float, Integer).

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	BulkByScrollTask.Status	Status of the reindex, update by query, or delete by query.
static class	BulkByScrollTask.StatusBuilder	This class acts as a builder for BulkByScrollTask.Status.
static class	BulkByScrollTask.StatusOrException	The status of a slice of the request.

Nested classes/interfaces inherited from class org.elasticsearch.tasks.CancellableTask

CancellableTask.CancellationListener

Field Summary

Fields inherited from class org.elasticsearch.tasks.Task

APM_TRACE_CONTEXT, HEADERS_TO_COPY, PARENT_APM_TRACE_CONTEXT, PARENT_TRACE_PARENT_HEADER, PARENT_TRACE_STATE, startTime, TRACE_ID, TRACE_PARENT_HTTP_HEADER, TRACE_START_TIME, TRACE_STATE, X_ELASTIC_PRODUCT_ORIGIN_HTTP_HEADER, X_ELASTIC_PROJECT_ID_HTTP_HEADER, X_OPAQUE_ID_HTTP_HEADER

Constructor Summary

Constructors

Constructor	Description
BulkByScrollTask(TaskId taskId, String type, String action, String description, TaskId parentTaskId, Map<String,String> headers, boolean eligibleForRelocationOnShutdown, ResumeInfo.RelocationOrigin relocationOrigin)

Method Summary

Modifier and Type	Method	Description
LeaderBulkByScrollTaskState	getLeaderState()	Returns the object that tracks the state of sliced subtasks.
BulkByScrollTask.Status	getStatus()	Build a status for this task or null if this task doesn't have status.
WorkerBulkByScrollTaskState	getWorkerState()	Returns the object that manages sending search requests.
boolean	isEligibleForRelocationOnShutdown()	Returns whether we should attempt to relocate this task to another node when the current node is preparing to shut down.
boolean	isLeader()	Returns true if this task is a leader for other slice subtasks
boolean	isRelocatedTask()	Returns true if this task was created via relocation from another node.
boolean	isRelocationRequested()	Returns whether this task has been marked as requiring relocation to another node.
boolean	isWorker()	Returns true if this task is a worker task that performs search requests.
void	onCancelled()	Called after the task is cancelled so that it can take any actions that it has to take.
ResumeInfo.RelocationOrigin	relocationOrigin()	Returns the relocation origin if this task is a relocated continuation.
void	requestRelocation()	Marks this task as requiring relocation to another node, e.g.
void	setWorker(float requestsPerSecond, Integer sliceId)	Sets this task to be a worker task that performs search requests
void	setWorkerCount(int slices)	Sets this task to be a leader task for slices sliced subtasks
TaskInfo	taskInfoGivenSubtaskInfo(String localNodeId, List<TaskInfo> sliceInfo)	Build the status for this task given a snapshot of the information of running slices.

Methods inherited from class org.elasticsearch.tasks.CancellableTask

addListener, ensureNotCancelled, getReasonCancelled, isCancelled, notifyIfCancelled, shouldCancelChildrenOnCancellation, toString

Methods inherited from class org.elasticsearch.tasks.Task

getAction, getDescription, getHeader, getId, getParentTaskId, getProjectId, getSpanId, getStartTime, getStartTimeNanos, getType, headers, result, result, taskInfo, taskInfo

Methods inherited from class java.lang.Object

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

Constructor Details

BulkByScrollTask

public BulkByScrollTask(TaskId taskId,
 String type,
 String action,
 String description,
 TaskId parentTaskId,
 Map<String,String> headers,
 boolean eligibleForRelocationOnShutdown,
 @Nullable
 ResumeInfo.RelocationOrigin relocationOrigin)

Method Details

getStatus

public BulkByScrollTask.Status getStatus()

Description copied from class: Task

Build a status for this task or null if this task doesn't have status. Since most tasks don't have status this defaults to returning null. While this can never perform IO it might be a costly operation, requiring collating lists of results, etc. So only use it if you need the value.

Overrides:

getStatus in class Task

taskInfoGivenSubtaskInfo

public TaskInfo taskInfoGivenSubtaskInfo(String localNodeId,
 List<TaskInfo> sliceInfo)

Build the status for this task given a snapshot of the information of running slices. This is only supported if the task is set as a leader for slice subtasks

isLeader

public boolean isLeader()

Returns true if this task is a leader for other slice subtasks

setWorkerCount

public void setWorkerCount(int slices)

Sets this task to be a leader task for slices sliced subtasks

getLeaderState

public LeaderBulkByScrollTaskState getLeaderState()

Returns the object that tracks the state of sliced subtasks. Throws IllegalStateException if this task is not set to be a leader task.

isWorker

public boolean isWorker()

Returns true if this task is a worker task that performs search requests. False otherwise

setWorker

public void setWorker(float requestsPerSecond,
 @Nullable
 Integer sliceId)

Sets this task to be a worker task that performs search requests

Parameters:

requestsPerSecond - How many search requests per second this task should make

sliceId - If this is is a sliced task, which slice number this task corresponds to. Null if not sliced.

getWorkerState

public WorkerBulkByScrollTaskState getWorkerState()

Returns the object that manages sending search requests. Throws IllegalStateException if this task is not set to be a worker task.

onCancelled

public void onCancelled()

Description copied from class: CancellableTask

Called after the task is cancelled so that it can take any actions that it has to take.

Overrides:

onCancelled in class CancellableTask

isEligibleForRelocationOnShutdown

public boolean isEligibleForRelocationOnShutdown()

Returns whether we should attempt to relocate this task to another node when the current node is preparing to shut down.

requestRelocation

public void requestRelocation()

Marks this task as requiring relocation to another node, e.g. because this node is about to shut down.

This method is fire-and-forget and does not guarantee that relocation actually happens. (That can depend on various factors, such as whether a suitable new node is available, and whether the task is able to get to an appropriate point to stop work and trigger the relocation in time.)

This should only be called on tasks where returns true.

isRelocationRequested

public boolean isRelocationRequested()

Returns whether this task has been marked as requiring relocation to another node. See requestRelocation().

relocationOrigin

public ResumeInfo.RelocationOrigin relocationOrigin()

Returns the relocation origin if this task is a relocated continuation.

isRelocatedTask

public boolean isRelocatedTask()

Returns true if this task was created via relocation from another node.