ClusterCoreDaemon

java.lang.Object
- akka.cluster.ClusterCoreDaemon

All Implemented Interfaces:: Actor, ActorLogging, RequiresMessageQueue<UnboundedMessageQueueSemantics>

public class ClusterCoreDaemon
extends java.lang.Object
implements Actor, ActorLogging, RequiresMessageQueue<UnboundedMessageQueueSemantics>

INTERNAL API.

Nested Class Summary

Nested Classes
Modifier and Type	Class and Description
`class`	`ClusterCoreDaemon.Ignored`
`class`	`ClusterCoreDaemon.Merge`
`class`	`ClusterCoreDaemon.Newer`
`class`	`ClusterCoreDaemon.Older`
`static interface`	`ClusterCoreDaemon.ReceiveGossipType` The types of gossip actions that receive gossip has performed.
`class`	`ClusterCoreDaemon.Same`

Nested classes/interfaces inherited from interface akka.actor.Actor
Actor.emptyBehavior$, Actor.ignoringBehavior$

Constructor Summary

Constructors
Constructor and Description

ClusterCoreDaemon(ActorRef publisher)

Constructors
Constructor and Description
`ClusterCoreDaemon(ActorRef publisher)`

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`double`	`adjustedGossipDifferentViewProbability()` For large clusters we should avoid shooting down individual nodes.
`void`	`assertLatestGossip()`
`void`	`becomeInitialized()`
`void`	`becomeUninitialized()`
`Cluster`	`cluster()`
`ActorContext`	`context()` Stores the context for this actor, including self, and sender.
`void`	`downing(Address address)` State transition to DOWN.
`Cancellable`	`failureDetectorReaperTask()`
`void`	`gossip()` Initiates a new round of gossip.
`void`	`gossipRandomN(int n)` Sends full gossip to `n` other random members.
`void`	`gossipSpeedupTick()`
`GossipStats`	`gossipStats()`
`void`	`gossipStatusTo(UniqueAddress node)`
`void`	`gossipStatusTo(UniqueAddress node, ActorRef destination)`
`Cancellable`	`gossipTask()`
`void`	`gossipTick()`
`void`	`gossipTo(UniqueAddress node)` Gossips latest gossip to a node.
`void`	`gossipTo(UniqueAddress node, ActorRef destination)`
`akka.cluster.ClusterCoreDaemon.Ignored$`	`Ignored()`
`scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit>`	`initialized()`
`void`	`initJoin()`
`boolean`	`isGossipSpeedupNeeded()`
`boolean`	`isMinNrOfMembersFulfilled()`
`boolean`	`isSingletonCluster()`
`void`	`join(Address address)` Try to join this cluster node with the node specified by `address`.
`void`	`joining(UniqueAddress node, scala.collection.immutable.Set<java.lang.String> roles)` State transition to JOINING - new node joining.
`void`	`joinSeedNodes(scala.collection.immutable.IndexedSeq<Address> newSeedNodes)`
`Gossip`	`latestGossip()`
`int`	`leaderActionCounter()`
`void`	`leaderActions()` Runs periodic leader actions, such as member status transitions, assigning partitions etc.
`void`	`leaderActionsOnConvergence()` Leader actions are as follows: 1.
`Cancellable`	`leaderActionsTask()`
`void`	`leaving(Address address)` State transition to LEAVING.
`int`	`MaxGossipsBeforeShuttingDownMyself()`
`akka.cluster.ClusterCoreDaemon.Merge$`	`Merge()`
`void`	`moveJoiningToWeaklyUp()`
`akka.cluster.ClusterCoreDaemon.Newer$`	`Newer()`
`int`	`NumberOfGossipsBeforeShutdownWhenLeaderExits()`
`akka.cluster.ClusterCoreDaemon.Older$`	`Older()`
`void`	`postStop()` User overridable callback.
`void`	`preStart()` User overridable callback.
`void`	`publish(Gossip newGossip)`
`void`	`publishInternalStats()`
`scala.Option<Cancellable>`	`publishStatsTask()`
`void`	`quarantined(UniqueAddress node)`
`void`	`reapUnreachableMembers()` Reaps the unreachable members according to the failure detector's verdict.
`scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit>`	`receive()` This defines the initial actor behavior, it must return a partial function with the actor logic.
`ClusterCoreDaemon.ReceiveGossipType`	`receiveGossip(GossipEnvelope envelope)` Receive new gossip.
`void`	`receiveGossipStatus(GossipStatus status)`
`scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit>`	`removed()`
`akka.cluster.ClusterCoreDaemon.Same$`	`Same()`
`scala.Option<ActorRef>`	`seedNodeProcess()`
`int`	`seedNodeProcessCounter()`
`scala.collection.immutable.IndexedSeq<Address>`	`seedNodes()`
`scala.Option<UniqueAddress>`	`selectRandomNode(scala.collection.IndexedSeq<UniqueAddress> nodes)`
`ActorRef`	`self()` The 'self' field holds the ActorRef for this actor.
`protected UniqueAddress`	`selfUniqueAddress()`
`void`	`sendGossipTo(Address address)`
`void`	`shutdown()` This method is called when a member sees itself as Exiting or Down.
`void`	`shutdownSelfWhenDown()`
`boolean`	`statsEnabled()`
`void`	`stopSeedNodeProcess()`
`scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit>`	`tryingToJoin(Address joinWith, scala.Option<scala.concurrent.duration.Deadline> deadline)`
`void`	`unhandled(java.lang.Object message)` User overridable callback.
`scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit>`	`uninitialized()`
`void`	`updateLatestGossip(Gossip newGossip)`
`boolean`	`validNodeForGossip(UniqueAddress node)`
`java.lang.String`	`vclockName(UniqueAddress node)`
`java.lang.String`	`vclockNode()`
`void`	`welcome(Address joinWith, UniqueAddress from, Gossip gossip)` Reply from Join request.

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface akka.actor.Actor
aroundPostRestart, aroundPostStop, aroundPreRestart, aroundPreStart, aroundReceive, postRestart, preRestart, sender, supervisorStrategy

Methods inherited from interface akka.actor.ActorLogging
log

- Constructor Detail
  - ClusterCoreDaemon
```
public ClusterCoreDaemon(ActorRef publisher)
```
- Method Detail
  - Ignored
```
public akka.cluster.ClusterCoreDaemon.Ignored$ Ignored()
```
  - Older
```
public akka.cluster.ClusterCoreDaemon.Older$ Older()
```
  - Newer
```
public akka.cluster.ClusterCoreDaemon.Newer$ Newer()
```
  - Same
```
public akka.cluster.ClusterCoreDaemon.Same$ Same()
```
  - Merge
```
public akka.cluster.ClusterCoreDaemon.Merge$ Merge()
```
  - context
```
public ActorContext context()
```
    Description copied from interface: Actor
    
    Stores the context for this actor, including self, and sender. It is implicit to support operations such as forward.
    WARNING: Only valid within the Actor itself, so do not close over it and publish it to other threads!
    ActorContext is the Scala API. getContext returns a UntypedActorContext, which is the Java API of the actor context.
    
    Specified by:
    
    context in interface Actor
    
    Returns:
    
    (undocumented)
  - self
```
public final ActorRef self()
```
    Description copied from interface: Actor
    The 'self' field holds the ActorRef for this actor.
    Can be used to send messages to itself:
```
 self ! message
 
```
    Specified by:
    
    self in interface Actor
    
    Returns:
    
    (undocumented)
  - cluster
```
public Cluster cluster()
```
  - selfUniqueAddress
```
protected UniqueAddress selfUniqueAddress()
```
  - NumberOfGossipsBeforeShutdownWhenLeaderExits
```
public int NumberOfGossipsBeforeShutdownWhenLeaderExits()
```
  - MaxGossipsBeforeShuttingDownMyself
```
public int MaxGossipsBeforeShuttingDownMyself()
```
  - vclockName
```
public java.lang.String vclockName(UniqueAddress node)
```
  - vclockNode
```
public java.lang.String vclockNode()
```
  - latestGossip
```
public Gossip latestGossip()
```
  - statsEnabled
```
public boolean statsEnabled()
```
  - gossipStats
```
public GossipStats gossipStats()
```
  - seedNodes
```
public scala.collection.immutable.IndexedSeq<Address> seedNodes()
```
  - seedNodeProcess
```
public scala.Option<ActorRef> seedNodeProcess()
```
  - seedNodeProcessCounter
```
public int seedNodeProcessCounter()
```
  - leaderActionCounter
```
public int leaderActionCounter()
```
  - gossipTask
```
public Cancellable gossipTask()
```
  - failureDetectorReaperTask
```
public Cancellable failureDetectorReaperTask()
```
  - leaderActionsTask
```
public Cancellable leaderActionsTask()
```
  - publishStatsTask
```
public scala.Option<Cancellable> publishStatsTask()
```
  - preStart
```
public void preStart()
```
    Description copied from interface: Actor
    
    User overridable callback.
    Is called when an Actor is started. Actors are automatically started asynchronously when created. Empty default implementation.
    
    Specified by:
    
    preStart in interface Actor
  - postStop
```
public void postStop()
```
    Description copied from interface: Actor
    
    User overridable callback.
    Is called asynchronously after 'actor.stop()' is invoked. Empty default implementation.
    
    Specified by:
    
    postStop in interface Actor
  - uninitialized
```
public scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit> uninitialized()
```
  - tryingToJoin
```
public scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit> tryingToJoin(Address joinWith,
                                                                                    scala.Option<scala.concurrent.duration.Deadline> deadline)
```
  - becomeUninitialized
```
public void becomeUninitialized()
```
  - becomeInitialized
```
public void becomeInitialized()
```
  - initialized
```
public scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit> initialized()
```
  - removed
```
public scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit> removed()
```
  - receive
```
public scala.PartialFunction<java.lang.Object,scala.runtime.BoxedUnit> receive()
```
    Description copied from interface: Actor
    
    This defines the initial actor behavior, it must return a partial function with the actor logic.
    
    Specified by:
    
    receive in interface Actor
    
    Returns:
    
    (undocumented)
  - unhandled
```
public void unhandled(java.lang.Object message)
```
    Description copied from interface: Actor
    
    User overridable callback.
    Is called when a message isn't handled by the current behavior of the actor by default it fails with either a DeathPactException (in case of an unhandled Terminated message) or publishes an UnhandledMessage to the actor's system's EventStream
    
    Specified by:
    
    unhandled in interface Actor
    
    Parameters:
    
    message - (undocumented)
  - initJoin
```
public void initJoin()
```
  - joinSeedNodes
```
public void joinSeedNodes(scala.collection.immutable.IndexedSeq<Address> newSeedNodes)
```
  - join
```
public void join(Address address)
```
    Try to join this cluster node with the node specified by address. It's only allowed to join from an empty state, i.e. when not already a member. A Join(selfUniqueAddress) command is sent to the node to join, which will reply with a Welcome message.
    
    Parameters:
    
    address - (undocumented)
  - stopSeedNodeProcess
```
public void stopSeedNodeProcess()
```
  - joining
```
public void joining(UniqueAddress node,
                    scala.collection.immutable.Set<java.lang.String> roles)
```
    State transition to JOINING - new node joining. Received Join message and replies with Welcome message, containing current gossip state, including the new joining member.
    
    Parameters:
    
    node - (undocumented)
    
    roles - (undocumented)
  - welcome
```
public void welcome(Address joinWith,
                    UniqueAddress from,
                    Gossip gossip)
```
    Reply from Join request.
    
    Parameters:
    
    joinWith - (undocumented)
    
    from - (undocumented)
    
    gossip - (undocumented)
  - leaving
```
public void leaving(Address address)
```
    State transition to LEAVING. The node will eventually be removed by the leader, after hand-off in EXITING, and only after removal a new node with same address can join the cluster through the normal joining procedure.
    
    Parameters:
    
    address - (undocumented)
  - shutdown
```
public void shutdown()
```
    This method is called when a member sees itself as Exiting or Down.
  - downing
```
public void downing(Address address)
```
    State transition to DOWN. Its status is set to DOWN. The node is also removed from the seen table.
    The node will eventually be removed by the leader, and only after removal a new node with same address can join the cluster through the normal joining procedure.
    
    Parameters:
    
    address - (undocumented)
  - quarantined
```
public void quarantined(UniqueAddress node)
```
  - receiveGossipStatus
```
public void receiveGossipStatus(GossipStatus status)
```
  - receiveGossip
```
public ClusterCoreDaemon.ReceiveGossipType receiveGossip(GossipEnvelope envelope)
```
    Receive new gossip.
    
    Parameters:
    
    envelope - (undocumented)
    
    Returns:
    
    (undocumented)
  - gossipTick
```
public void gossipTick()
```
  - gossipSpeedupTick
```
public void gossipSpeedupTick()
```
  - isGossipSpeedupNeeded
```
public boolean isGossipSpeedupNeeded()
```
  - gossipRandomN
```
public void gossipRandomN(int n)
```
    Sends full gossip to n other random members.
    
    Parameters:
    
    n - (undocumented)
  - gossip
```
public void gossip()
```
    Initiates a new round of gossip.
  - adjustedGossipDifferentViewProbability
```
public double adjustedGossipDifferentViewProbability()
```
    For large clusters we should avoid shooting down individual nodes. Therefore the probability is reduced for large clusters.
    
    Returns:
    
    (undocumented)
  - leaderActions
```
public void leaderActions()
```
    Runs periodic leader actions, such as member status transitions, assigning partitions etc.
  - shutdownSelfWhenDown
```
public void shutdownSelfWhenDown()
```
  - isMinNrOfMembersFulfilled
```
public boolean isMinNrOfMembersFulfilled()
```
  - leaderActionsOnConvergence
```
public void leaderActionsOnConvergence()
```
    Leader actions are as follows: 1. Move JOINING => UP -- When a node joins the cluster 2. Move LEAVING => EXITING -- 3. Non-exiting remain -- 4. Move unreachable EXITING => REMOVED -- When all nodes have seen the EXITING node as unreachable (convergence) - remove the node from the node ring and seen table 5. Move unreachable DOWN/EXITING => REMOVED -- When all nodes have seen that the node is DOWN/EXITING (convergence) - remove the node from the node ring and seen table 7. Updating the vclock version for the changes 8. Updating the seen table 9. Update the state with the new gossip
  - moveJoiningToWeaklyUp
```
public void moveJoiningToWeaklyUp()
```
  - reapUnreachableMembers
```
public void reapUnreachableMembers()
```
    Reaps the unreachable members according to the failure detector's verdict.
  - selectRandomNode
```
public scala.Option<UniqueAddress> selectRandomNode(scala.collection.IndexedSeq<UniqueAddress> nodes)
```
  - isSingletonCluster
```
public boolean isSingletonCluster()
```
  - sendGossipTo
```
public void sendGossipTo(Address address)
```
  - gossipTo
```
public void gossipTo(UniqueAddress node)
```
    Gossips latest gossip to a node.
    
    Parameters:
    
    node - (undocumented)
  - gossipTo
```
public void gossipTo(UniqueAddress node,
                     ActorRef destination)
```
  - gossipStatusTo
```
public void gossipStatusTo(UniqueAddress node,
                           ActorRef destination)
```
  - gossipStatusTo
```
public void gossipStatusTo(UniqueAddress node)
```
  - validNodeForGossip
```
public boolean validNodeForGossip(UniqueAddress node)
```
  - updateLatestGossip
```
public void updateLatestGossip(Gossip newGossip)
```
  - assertLatestGossip
```
public void assertLatestGossip()
```
  - publish
```
public void publish(Gossip newGossip)
```
  - publishInternalStats
```
public void publishInternalStats()
```

Class ClusterCoreDaemon

Nested Class Summary

Nested classes/interfaces inherited from interface akka.actor.Actor

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Methods inherited from interface akka.actor.Actor

Methods inherited from interface akka.actor.ActorLogging

Constructor Detail

ClusterCoreDaemon

Method Detail

Ignored

Older

Newer

Same

Merge

context

self

cluster

selfUniqueAddress

NumberOfGossipsBeforeShutdownWhenLeaderExits

MaxGossipsBeforeShuttingDownMyself

vclockName

vclockNode

latestGossip

statsEnabled

gossipStats

seedNodes

seedNodeProcess

seedNodeProcessCounter

leaderActionCounter

gossipTask

failureDetectorReaperTask

leaderActionsTask

publishStatsTask

preStart

postStop

uninitialized

tryingToJoin

becomeUninitialized

becomeInitialized

initialized

removed

receive

unhandled

initJoin

joinSeedNodes

join

stopSeedNodeProcess

joining

welcome

leaving

shutdown

downing

quarantined

receiveGossipStatus

receiveGossip

gossipTick

gossipSpeedupTick

isGossipSpeedupNeeded

gossipRandomN

gossip

adjustedGossipDifferentViewProbability

leaderActions

shutdownSelfWhenDown

isMinNrOfMembersFulfilled

leaderActionsOnConvergence

moveJoiningToWeaklyUp

reapUnreachableMembers

selectRandomNode

isSingletonCluster

sendGossipTo

gossipTo

gossipTo

gossipStatusTo

gossipStatusTo

validNodeForGossip

updateLatestGossip

assertLatestGossip

publish