Package org.elasticsearch.cluster.action.shard

Class ShardStateAction

java.lang.Object

org.elasticsearch.cluster.action.shard.ShardStateAction

public class ShardStateAction
extends Object

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	ShardStateAction.FailedShardEntry
static class	ShardStateAction.NoLongerPrimaryShardException
static class	ShardStateAction.ShardFailedClusterStateTaskExecutor
static class	ShardStateAction.ShardStartedClusterStateTaskExecutor
static class	ShardStateAction.StartedShardEntry

Field Summary

Fields

Modifier and Type	Field	Description
static Setting<Priority>	FOLLOW_UP_REROUTE_PRIORITY_SETTING	Adjusts the priority of the followup reroute task.
static String	SHARD_FAILED_ACTION_NAME
static String	SHARD_STARTED_ACTION_NAME

Constructor Summary

Constructors

Constructor	Description
ShardStateAction(ClusterService clusterService, TransportService transportService, AllocationService allocationService, RerouteService rerouteService, ThreadPool threadPool)

Method Summary

Modifier and Type	Method	Description
void	localShardFailed(ShardRouting shardRouting, String message, Exception failure, ActionListener<Void> listener)	Send a shard failed request to the master node to update the cluster state when a shard on the local node failed.
void	localShardFailed(ShardRouting shardRouting, String message, Exception failure, ActionListener<Void> listener, ClusterState currentState)	Send a shard failed request to the master node to update the cluster state when a shard on the local node failed.
void	remoteShardFailed(ShardId shardId, String allocationId, long primaryTerm, boolean markAsStale, String message, Exception failure, ActionListener<Void> listener)	Send a shard failed request to the master node to update the cluster state with the failure of a shard on another node.
void	shardStarted(ShardRouting shardRouting, long primaryTerm, String message, ShardLongFieldRange timestampRange, ActionListener<Void> listener)
void	shardStarted(ShardRouting shardRouting, long primaryTerm, String message, ShardLongFieldRange timestampRange, ActionListener<Void> listener, ClusterState currentState)
protected void	waitForNewMasterAndRetry(String actionName, ClusterStateObserver observer, TransportRequest request, ActionListener<Void> listener, Predicate<ClusterState> changePredicate)

Methods inherited from class java.lang.Object

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

Field Details

SHARD_STARTED_ACTION_NAME

public static final String SHARD_STARTED_ACTION_NAME

See Also:

Constant Field Values

SHARD_FAILED_ACTION_NAME

public static final String SHARD_FAILED_ACTION_NAME

See Also:

Constant Field Values

FOLLOW_UP_REROUTE_PRIORITY_SETTING

public static final Setting<Priority> FOLLOW_UP_REROUTE_PRIORITY_SETTING

Adjusts the priority of the followup reroute task. NORMAL is right for reasonable clusters, but in a badly configured cluster it may be necessary to raise this higher to recover the older behaviour of rerouting after processing every shard-started task. Deliberately undocumented, since this is a last-resort escape hatch for experts rather than something we want to expose to anyone, and deprecated since we will remove it once we have confirmed from experience that this priority is appropriate in all cases.

Constructor Details

ShardStateAction

@Inject
public ShardStateAction(ClusterService clusterService,
 TransportService transportService,
 AllocationService allocationService,
 RerouteService rerouteService,
 ThreadPool threadPool)

Method Details

remoteShardFailed

public void remoteShardFailed(ShardId shardId,
 String allocationId,
 long primaryTerm,
 boolean markAsStale,
 String message,
 @Nullable
 Exception failure,
 ActionListener<Void> listener)

Send a shard failed request to the master node to update the cluster state with the failure of a shard on another node. This means that the shard should be failed because a write made it into the primary but was not replicated to this shard copy. If the shard does not exist anymore but still has an entry in the in-sync set, remove its allocation id from the in-sync set.

Parameters:

shardId - shard id of the shard to fail

allocationId - allocation id of the shard to fail

primaryTerm - the primary term associated with the primary shard that is failing the shard. Must be strictly positive.

markAsStale - whether or not to mark a failing shard as stale (eg. removing from in-sync set) when failing the shard.

message - the reason for the failure

failure - the underlying cause of the failure

listener - callback upon completion of the request

localShardFailed

public void localShardFailed(ShardRouting shardRouting,
 String message,
 @Nullable
 Exception failure,
 ActionListener<Void> listener)

Send a shard failed request to the master node to update the cluster state when a shard on the local node failed.

localShardFailed

public void localShardFailed(ShardRouting shardRouting,
 String message,
 @Nullable
 Exception failure,
 ActionListener<Void> listener,
 ClusterState currentState)

Send a shard failed request to the master node to update the cluster state when a shard on the local node failed.

waitForNewMasterAndRetry

protected void waitForNewMasterAndRetry(String actionName,
 ClusterStateObserver observer,
 TransportRequest request,
 ActionListener<Void> listener,
 Predicate<ClusterState> changePredicate)

shardStarted

public void shardStarted(ShardRouting shardRouting,
 long primaryTerm,
 String message,
 ShardLongFieldRange timestampRange,
 ActionListener<Void> listener)

shardStarted

public void shardStarted(ShardRouting shardRouting,
 long primaryTerm,
 String message,
 ShardLongFieldRange timestampRange,
 ActionListener<Void> listener,
 ClusterState currentState)