com.hazelcast.internal.cluster.fd

Class PhiAccrualFailureDetector

java.lang.Object

com.hazelcast.internal.cluster.fd.PhiAccrualFailureDetector

All Implemented Interfaces:

FailureDetector

public class PhiAccrualFailureDetector
extends Object
implements FailureDetector

Port of Akka's PhiAccrualFailureDetector.scala

Implementation of 'The Phi Accrual Failure Detector' by Hayashibara et al. as defined in their paper.

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:

 φ = -log10(1 - F(timeSinceLastHeartbeat)
 

where F is the cumulative distribution function of a normal distribution with mean and standard deviation estimated from historical heartbeat inter-arrival times.

Constructor Summary

Constructors

Constructor and Description
PhiAccrualFailureDetector(double threshold, int maxSampleSize, double minStdDeviationMillis, long acceptableHeartbeatPauseMillis, long firstHeartbeatEstimateMillis)

Method Summary

Modifier and Type	Method and Description
void	heartbeat(long timestampMillis) Notifies this failure detector about received heartbeat message from the tracked member.
boolean	isAlive(long timestampMillis) Returns true if the tracked member is considered as alive/available.
long	lastHeartbeat() Returns the last heartbeat timestamp for the tracked member.
double	suspicionLevel(long timestamp) Returns suspicion level about the tracked member.

Methods inherited from class java.lang.Object

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

Constructor Detail

PhiAccrualFailureDetector

public PhiAccrualFailureDetector(double threshold,
                                 int maxSampleSize,
                                 double minStdDeviationMillis,
                                 long acceptableHeartbeatPauseMillis,
                                 long firstHeartbeatEstimateMillis)

Parameters:

threshold - A low threshold is prone to generate many wrong suspicions but ensures a quick detection in the event of a real crash. Conversely, a high threshold generates fewer mistakes but needs more time to detect actual crashes

maxSampleSize - Number of samples to use for calculation of mean and standard deviation of inter-arrival times.

minStdDeviationMillis - Minimum standard deviation to use for the normal distribution used when calculating phi. Too low standard deviation might result in too much sensitivity for sudden, but normal, deviations in heartbeat inter arrival times.

acceptableHeartbeatPauseMillis - Duration corresponding to number of potentially lost/delayed heartbeats that will be accepted before considering it to be an anomaly. This margin is important to be able to survive sudden, occasional, pauses in heartbeat arrivals, due to for example garbage collect or network drop.

firstHeartbeatEstimateMillis - Bootstrap the stats with heartbeats that corresponds to this duration, with a with rather high standard deviation (since environment is unknown in the beginning)

Method Detail

isAlive

public boolean isAlive(long timestampMillis)

Description copied from interface: FailureDetector

Returns true if the tracked member is considered as alive/available.

Specified by:

isAlive in interface FailureDetector

Parameters:

timestampMillis - timestamp in milliseconds

Returns:

true if the member is alive

heartbeat

public void heartbeat(long timestampMillis)

Description copied from interface: FailureDetector

Notifies this failure detector about received heartbeat message from the tracked member.

Specified by:

heartbeat in interface FailureDetector

Parameters:

timestampMillis - timestamp of heartbeat message in milliseconds

lastHeartbeat

public long lastHeartbeat()

Description copied from interface: FailureDetector

Returns the last heartbeat timestamp for the tracked member.

Specified by:

lastHeartbeat in interface FailureDetector

Returns:

heartbeat timestamp in milliseconds

suspicionLevel

public double suspicionLevel(long timestamp)

Description copied from interface: FailureDetector

Returns suspicion level about the tracked member. Returned value is mostly implementation dependent. 0 indicates no suspicion at all.

Specified by:

suspicionLevel in interface FailureDetector

Parameters:

timestamp - timestamp in milliseconds

Returns:

suspicion level