Package org.exoplatform.commons.search.index.impl

Class ElasticIndexingOperationProcessor

java.lang.Object

org.exoplatform.commons.search.index.IndexingOperationProcessor

org.exoplatform.commons.search.index.impl.ElasticIndexingOperationProcessor

All Implemented Interfaces:

org.picocontainer.Startable

public class ElasticIndexingOperationProcessor
extends IndexingOperationProcessor
implements org.picocontainer.Startable

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
class	ElasticIndexingOperationProcessor.ReindexESType

Constructor Summary

Constructors

Constructor	Description
ElasticIndexingOperationProcessor(IndexingOperationDAO indexingOperationDAO, ElasticIndexingClient elasticIndexingClient, ElasticContentRequestBuilder elasticContentRequestBuilder, ElasticIndexingAuditTrail auditTrail, org.exoplatform.commons.persistence.impl.EntityManagerService entityManagerService, org.exoplatform.container.xml.InitParams initParams)

Method Summary

Modifier and Type	Method	Description
void	addConnector(IndexingServiceConnector indexingServiceConnector)	Add Indexing Connector to the service
void	addConnector(IndexingServiceConnector indexingServiceConnector, Boolean override)	Add Indexing Connector to the service
Integer	getBatchNumber()
Integer	getRequestSizeLimit()
void	interrupt()	Set the indexing process as interrupted in order to terminate it as soon as possible without finishing the whole process.
void	process()	Handle the Indexing queue Get all data in the indexing queue, transform them to ES requests, send requests to ES This method is ONLY called by the job scheduler.
void	setBatchNumber(Integer batchNumber)
void	setInitialized(boolean initialized)
void	setRequestSizeLimit(Integer requestSizeLimit)
void	start()
void	stop()

Methods inherited from class org.exoplatform.commons.search.index.IndexingOperationProcessor

getConnectors

Methods inherited from class java.lang.Object

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

Constructor Details

ElasticIndexingOperationProcessor

public ElasticIndexingOperationProcessor(IndexingOperationDAO indexingOperationDAO,
 ElasticIndexingClient elasticIndexingClient,
 ElasticContentRequestBuilder elasticContentRequestBuilder,
 ElasticIndexingAuditTrail auditTrail,
 org.exoplatform.commons.persistence.impl.EntityManagerService entityManagerService,
 org.exoplatform.container.xml.InitParams initParams)

Method Details

addConnector

public void addConnector(IndexingServiceConnector indexingServiceConnector)

Description copied from class: IndexingOperationProcessor

Add Indexing Connector to the service

Overrides:

addConnector in class IndexingOperationProcessor

Parameters:

indexingServiceConnector - the indexing connector to add

addConnector

public void addConnector(IndexingServiceConnector indexingServiceConnector,
 Boolean override)

Description copied from class: IndexingOperationProcessor

Add Indexing Connector to the service

Overrides:

addConnector in class IndexingOperationProcessor

Parameters:

indexingServiceConnector - the indexing connector to add

override - equal true if we can override an existing connector, false otherwise

process

public void process()

Handle the Indexing queue Get all data in the indexing queue, transform them to ES requests, send requests to ES This method is ONLY called by the job scheduler. This method is not annotated with @ExoTransactional because we don't want it to be executed in one transaction. A request lifecycle is started and ended for all jobs, it is done by org.exoplatform.services.scheduler.impl.JobEnvironmentConfigListener. It means that we have 1 entity manager per job execution. Because of that, we have to take care of cleaning the persistence context regularly to avoid to have too big sessions and bad performances. This method is synchronized to make sure the queue is processed by only one thread at a time, since the indexing queue does not support multi-thread processing for the moment.

Specified by:

process in class IndexingOperationProcessor

interrupt

public void interrupt()

Set the indexing process as interrupted in order to terminate it as soon as possible without finishing the whole process. Since the indexing process can take time (for a reindexAll operation for example), it allows to interrupt it gracefully (without killing the thread).

Specified by:

interrupt in class IndexingOperationProcessor

getBatchNumber

public Integer getBatchNumber()

setBatchNumber

public void setBatchNumber(Integer batchNumber)

getRequestSizeLimit

public Integer getRequestSizeLimit()

setRequestSizeLimit

public void setRequestSizeLimit(Integer requestSizeLimit)

start

public void start()

Specified by:

start in interface org.picocontainer.Startable

stop

public void stop()

Specified by:

stop in interface org.picocontainer.Startable

setInitialized

public void setInitialized(boolean initialized)