* Removed need for extra actor that subscribes to cluster
events
* ClusterRouterActor is the router actor and aslo subscribes
* Less scary initialization, and safe calls to createRoutees
* Intercept ClusterDomainEvent to route them to ClusterRouterActor
instead of to the routees
* Ordinary withRouther withDeploy is TOO ugly for this
* Implicit conversion adds withClusterRouter to Props
for Scala API
* Helper method in ClusterRouterPropsDecorator for Java API
* Config max-nr-of-instances-per-node
* selectDeploymentTarget that takes max-nr-of-instances-per-node
and nr-of-instances into account
* Deploy when new member added or removed
* Moved routeeProps to RouteeProvider constructor, needed for
this feature, but also simplifies createRoute, createRoutee,
and resize, since routeeProps doesn't have to be passed around.
* We will write more tests that rely on real Cluster(system) extension,
such as ClusterRoundRobinRoutedActorSpec
* When not using FailureDetectorStrategy or overriding seed nodes
MultiNodeClusterSpec will use the real Cluster(system) extension
instead of a new Cluster instance with additional test facilities
* Defined the domain events in ClusterEvent.scala file
* Produce events from diff and publish publish to event bus
from separate actor, ClusterDomainEventPublisher
* Adjustments of tests
* Gossip is not exposed in user api
* Better and more events
* Snapshot event sent to new subscriber
* Updated tests
* Periodic publish only for internal stats
* Trying to simultaneously resolving conflicts at several nodes creates new conflicts.
Therefore the leader resolves conflicts to limit divergence. To avoid overload there
is also a configurable rate limit of how many conflicts that are handled by second.
* Netty blocks when sending to broken connections. ClusterHeartbeatSender actor
isolates sending to different nodes by using child workers for each target
address and thereby reduce the risk of irregular heartbeats to healty
nodes due to broken connections to other nodes.
* Essentially as already described in cluster specification,
but now fully implemented and tested with LargeClusterSpec
* Gossip to nodes with different view (using seen table)
with certain probability
* Gossip chat, gossip back to sender
* Immediate gossip to joining node
* Updated some tests to reflect current implementation
* Add new operators :+ and :++ by implicit conversion
* Unfortunately this means that we must remember to use
these until SI-5986 is fixed. Is there a better way?
* self is initially not member (in gossip state)
* if the join to seed nodes timeout it joins itself, and becomes
singleton cluster
* remove the special case handling of singelton cluster in gossip
merge, since singleton cluster is not the normal state when joining
any more
