org.elasticsearch.index.shard

Class ShadowIndexShard

java.lang.Object

org.elasticsearch.index.shard.AbstractIndexShardComponent

org.elasticsearch.index.shard.IndexShard

org.elasticsearch.index.shard.ShadowIndexShard

All Implemented Interfaces:

IndexShardComponent, IndicesClusterStateService.Shard

public final class ShadowIndexShard
extends IndexShard

ShadowIndexShard extends IndexShard to add file synchronization from the primary when a flush happens. It also ensures that a replica being promoted to a primary causes the shard to fail, kicking off a re-allocation of the primary shard.

Nested Class Summary

Nested classes/interfaces inherited from class org.elasticsearch.index.shard.IndexShard

IndexShard.ShardFailure

Field Summary

Fields inherited from class org.elasticsearch.index.shard.IndexShard

currentEngineReference, engineFactory, primaryTerm, shardRouting, state, writeAllowedStatesForPrimary

Fields inherited from class org.elasticsearch.index.shard.AbstractIndexShardComponent

deprecationLogger, indexSettings, logger, shardId

Constructor Summary

Constructors

Constructor and Description

ShadowIndexShard(ShardRouting shardRouting, IndexSettings indexSettings, ShardPath path, Store store, IndexCache indexCache, MapperService mapperService, SimilarityService similarityService, IndexFieldDataService indexFieldDataService, EngineFactory engineFactory, IndexEventListener indexEventListener, IndexSearcherWrapper wrapper, ThreadPool threadPool, BigArrays bigArrays, Engine.Warmer engineWarmer, java.util.List<SearchOperationListener> searchOperationListeners)

Method Summary

Modifier and Type	Method and Description
void	addRefreshListener(Translog.Location location, java.util.function.Consumer<java.lang.Boolean> listener) Add a listener for refreshes.
protected RefreshListeners	buildRefreshListeners() Build RefreshListeners for this shard.
boolean	canIndex() returns true if this shard supports indexing (i.e., write) operations.
MergeStats	mergeStats()
protected Engine	newEngine(EngineConfig config)
protected void	onNewEngine(Engine newEngine)
boolean	shouldFlush() Returns true iff this shard needs to be flushed due to too many translog operation or a too large transaction log.
Store.MetadataSnapshot	snapshotStoreMetadata() gets a Store.MetadataSnapshot for the current directory.
TranslogStats	translogStats()
void	updateRoutingEntry(ShardRouting newRouting) In addition to the regular accounting done in IndexShard.updateRoutingEntry(ShardRouting), if this shadow replica needs to be promoted to a primary, the shard is failed in order to allow a new primary to be re-allocated.

Methods inherited from class org.elasticsearch.index.shard.IndexShard

acquireIndexCommit, acquirePrimaryOperationLock, acquireReplicaOperationLock, acquireSearcher, acquireTranslogView, activateThrottling, addShardFailureCallback, checkIdle, close, commitStats, completionStats, deactivateThrottling, delete, docStats, failShard, fieldData, fieldDataStats, finalizeRecovery, flush, flushAndCloseEngine, flushStats, forceMerge, get, getActiveOperationsCount, getEngineOrNull, getIndexBufferRAMBytesUsed, getIndexEventListener, getPrimaryTerm, getQueryCachingPolicy, getSearchOperationListener, getService, getStats, getTranslog, getTranslogDurability, getWritingBytes, ignoreRecoveryAttempt, index, indexFieldDataService, indexingStats, isActive, isRefreshNeeded, mapperService, markAsRecovering, maybeFlush, minimumCompatibleVersion, noopUpdate, onSettingsChanged, performBatchRecovery, performRecoveryRestart, performTranslogRecovery, postRecovery, prepareDeleteOnPrimary, prepareDeleteOnReplica, prepareForIndexRecovery, prepareIndexOnPrimary, prepareIndexOnReplica, readAllowed, recoverFromLocalShards, recoverFromStore, recoveryState, recoveryStats, refresh, refreshStats, releaseIndexCommit, relocated, requestCache, restoreFromRepository, routingEntry, searchStats, segments, segmentStats, shardBitsetFilterCache, shardPath, skipTranslogRecovery, startRecovery, state, store, storeStats, sync, syncFlush, updatePrimaryTerm, upgrade, verifyActive, warmerService, warmerStats, writeIndexingBuffer

Methods inherited from class org.elasticsearch.index.shard.AbstractIndexShardComponent

indexSettings, nodeName, shardId

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.elasticsearch.indices.cluster.IndicesClusterStateService.Shard

shardId

Constructor Detail

ShadowIndexShard

public ShadowIndexShard(ShardRouting shardRouting,
                        IndexSettings indexSettings,
                        ShardPath path,
                        Store store,
                        IndexCache indexCache,
                        MapperService mapperService,
                        SimilarityService similarityService,
                        IndexFieldDataService indexFieldDataService,
                        @Nullable
                        EngineFactory engineFactory,
                        IndexEventListener indexEventListener,
                        IndexSearcherWrapper wrapper,
                        ThreadPool threadPool,
                        BigArrays bigArrays,
                        Engine.Warmer engineWarmer,
                        java.util.List<SearchOperationListener> searchOperationListeners)
                 throws java.io.IOException

Throws:

java.io.IOException

Method Detail

updateRoutingEntry

public void updateRoutingEntry(ShardRouting newRouting)
                        throws java.io.IOException

In addition to the regular accounting done in IndexShard.updateRoutingEntry(ShardRouting), if this shadow replica needs to be promoted to a primary, the shard is failed in order to allow a new primary to be re-allocated.

Specified by:

updateRoutingEntry in interface IndicesClusterStateService.Shard

Overrides:

updateRoutingEntry in class IndexShard

Throws:

java.io.IOException - if shard state could not be persisted

mergeStats

public MergeStats mergeStats()

Overrides:

mergeStats in class IndexShard

canIndex

public boolean canIndex()

Description copied from class: IndexShard

returns true if this shard supports indexing (i.e., write) operations.

Overrides:

canIndex in class IndexShard

newEngine

protected Engine newEngine(EngineConfig config)

Overrides:

newEngine in class IndexShard

buildRefreshListeners

protected RefreshListeners buildRefreshListeners()

Description copied from class: IndexShard

Build RefreshListeners for this shard. Protected so ShadowIndexShard can override it to return null.

Overrides:

buildRefreshListeners in class IndexShard

shouldFlush

public boolean shouldFlush()

Description copied from class: IndexShard

Returns true iff this shard needs to be flushed due to too many translog operation or a too large transaction log. Otherwise false.

translogStats

public TranslogStats translogStats()

Overrides:

translogStats in class IndexShard

addRefreshListener

public void addRefreshListener(Translog.Location location,
                               java.util.function.Consumer<java.lang.Boolean> listener)

Description copied from class: IndexShard

Add a listener for refreshes.

Overrides:

addRefreshListener in class IndexShard

Parameters:

location - the location to listen for

listener - for the refresh. Called with true if registering the listener ran it out of slots and forced a refresh. Called with false otherwise.

snapshotStoreMetadata

public Store.MetadataSnapshot snapshotStoreMetadata()
                                             throws java.io.IOException

Description copied from class: IndexShard

gets a Store.MetadataSnapshot for the current directory. This method is safe to call in all lifecycle of the index shard, without having to worry about the current state of the engine and concurrent flushes.

Overrides:

snapshotStoreMetadata in class IndexShard

Throws:

org.apache.lucene.index.IndexNotFoundException - if no index is found in the current directory

org.apache.lucene.index.CorruptIndexException - if the lucene index is corrupted. This can be caused by a checksum mismatch or an unexpected exception when opening the index reading the segments file.

org.apache.lucene.index.IndexFormatTooOldException - if the lucene index is too old to be opened.

org.apache.lucene.index.IndexFormatTooNewException - if the lucene index is too new to be opened.

java.io.FileNotFoundException - if one or more files referenced by a commit are not present.

java.nio.file.NoSuchFileException - if one or more files referenced by a commit are not present.

java.io.IOException

onNewEngine

protected void onNewEngine(Engine newEngine)

Overrides:

onNewEngine in class IndexShard