Package com.yahoo.vespa.model.content.engines

Class ProtonEngine.Factory

java.lang.Object

com.yahoo.vespa.model.content.engines.ProtonEngine.Factory

All Implemented Interfaces:

PersistenceEngine.PersistenceFactory

Enclosing class:

ProtonEngine

public static class ProtonEngine.Factory
extends Object
implements PersistenceEngine.PersistenceFactory

Constructor Summary

Constructors

Constructor	Description
Factory(ContentSearchCluster search)

Method Summary

Modifier and Type	Method	Description
PersistenceEngine	create(DeployState deployState, StorageNode storageNode, StorageGroup parentGroup, ModelElement storageNodeElement)
boolean	enableMultiLevelSplitting()	Multi level splitting can increase split performance a lot where documents have been co-localized, for backends where retrieving document identifiers contained in bucket is cheap.
ContentCluster.DistributionMode	getDefaultDistributionMode()
boolean	supportRevert()	If a write request succeeds on some nodes and fails on others, causing request to fail to client, the content layer will revert the operation where it succeeded if reverts are supported.

Methods inherited from class java.lang.Object

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

Constructor Details

Factory

public Factory(ContentSearchCluster search)

Method Details

create

public PersistenceEngine create(DeployState deployState,
 StorageNode storageNode,
 StorageGroup parentGroup,
 ModelElement storageNodeElement)

Specified by:

create in interface PersistenceEngine.PersistenceFactory

supportRevert

public boolean supportRevert()

Description copied from interface: PersistenceEngine.PersistenceFactory

If a write request succeeds on some nodes and fails on others, causing request to fail to client, the content layer will revert the operation where it succeeded if reverts are supported. (Typically require backend to keep multiple entries of the same document identifier persisted at the same time)

Specified by:

supportRevert in interface PersistenceEngine.PersistenceFactory

enableMultiLevelSplitting

public boolean enableMultiLevelSplitting()

Description copied from interface: PersistenceEngine.PersistenceFactory

Multi level splitting can increase split performance a lot where documents have been co-localized, for backends where retrieving document identifiers contained in bucket is cheap. Backends where split is cheaper than fetching document identifiers will not want to enable multi level splitting.

Specified by:

enableMultiLevelSplitting in interface PersistenceEngine.PersistenceFactory

getDefaultDistributionMode

public ContentCluster.DistributionMode getDefaultDistributionMode()

Specified by:

getDefaultDistributionMode in interface PersistenceEngine.PersistenceFactory