Extension that holds a UID that is assigned as a random 'Int'. The UID is intended to be used together with an Address to be able to distinguish restarted actor system using the same host and port.
This class represents the latest date or time by which an operation should be completed.
This class represents a FailureDetector that uses an absolute timeout of missing heartbeats to trigger unavailability.
A lock-less, thread-safe implementation of IFailureDetectorRegistryT.
A failure detector must be a thread-safe, mutable construct that registers heartbeat events of a resource and is able to decide the availability of that monitored resource
INTERNAL API Utility class to create FailureDetector instances via reflection.
Implementation of 'The Phi Accrual Failure Detector' by Hayashibara et al. as defined in their paper: [http://ddg.jaist.ac.jp/pub/HDY+04.pdf] The suspicion level of failure is given by a value called φ (phi). The basic idea of the φ failure detector is to express the value of φ on a scale that is dynamically adjusted to reflect current network conditions. A configurable threshold is used to decide if φ is considered to be a failure. The value of φ is calculated as:
where F is the cumulative distribution function of a normal distribution with mean and standard deviation estimated from historical heartbeat inter-arrival times.
φ = -log10(1 - F(timeSinceLastHeartbeat)
This class represents configuration information used when setting up remoting.
INTERNAL API. The remote transport is responsible for sending and receiving messages. Each transport has an address, which it should provide in Serialization.CurrentTransportInformation (thread-local) while serializing ActorReferences (which might also be part of messages). This address must be available (i.e. fully initialized) by the time the first message is received or when the Start() method returns, whichever happens first.
This exception is thrown when a general failure within a RemoteTransport occurs, such as the inability to start, wrong configuration, etc...
INTERNAL API Remote nodes with actors that are watched are monitored by this actor to be able to detect network failures and process crashes. RemoteActorRefProvider intercepts Watch and Unwatch system messages and sends corresponding RemoteWatcherWatchRemote and RemoteWatcherUnwatchRemote to this actor. For a new node to be watched this actor periodically sends RemoteWatcherHeartbeat to the peer actor on the other node, which replies with RemoteWatcherHeartbeatRsp message back. The failure detector on the watching side monitors these heartbeat messages. If arrival of heartbeat messages stops it will be detected and this actor will publish to the . When all actors on a node have been unwatched it will stop sending heartbeat messages. For bi-directional watch between two nodes the same thing will be established in both directions, but independent of each other.
Remote lifecycle events that are published to the EventStream when initialization / connect / disconnect events that occur during network operations
Interface for a registry of Akka FailureDetectors. New resources are implicitly registered when heartbeat is first called with the resource given as parameter.
Abstraction of a clock that returns time in milliseconds. Can only be used to measure the elapsed time and is not related to any other notion of system or wall-clock time.