Package io.hekate.cluster.health

Class DefaultFailureDetector

java.lang.Object

io.hekate.cluster.health.DefaultFailureDetector

All Implemented Interfaces:

FailureDetector, JmxSupport<DefaultFailureDetectorJmx>, ConfigReportSupport

public class DefaultFailureDetector
extends Object
implements FailureDetector, JmxSupport<DefaultFailureDetectorJmx>, ConfigReportSupport

Default implementation of FailureDetector interface.

This implementation organizes all nodes into a ring with each node being monitored by a configurable number of other nodes. Monitoring is based on periodic exchange of heartbeat messages with a configurable time interval and a configurable amount of acceptable heartbeat losses.

Configuration options of this failure detector are represented by the DefaultFailureDetectorConfig class with the following properties:

• Failure detection quorum - specifies how many nodes will be monitored by this node
• Heartbeat interval - time interval in milliseconds between attempts of sending a heartbeat request message to the monitored node
• Heartbeat loss threshold - acceptable amount of lost heartbeats before considering the monitored node to be failed. Note that heartbeat is considered to be lost if no response for it was received during the heartbeat interval.

See Also:

DefaultFailureDetectorConfig, ClusterServiceFactory.setFailureDetector(FailureDetector)

Constructor Summary

Constructors

Constructor	Description
DefaultFailureDetector()	Constructs a new instance with the default configuration options.
DefaultFailureDetector(DefaultFailureDetectorConfig cfg)	Constructs a new instance using the specified configuration.

Method Summary

Modifier and Type	Method	Description
int	failureQuorum()	Return the amount of nodes that should agree on some particular node failure before removing such node from the cluster.
long	heartbeatInterval()	Returns the time interval in milliseconds between heartbeat sending rounds (see FailureDetector.heartbeatTick()).
int	heartbeatLossThreshold()	Returns the amount of heartbeats that can be lost before considering a node failure.
Collection<ClusterAddress>	heartbeatTick()	Runs a heartbeat tick and returns a set of cluster node addresses that should received a heartbeat request message.
void	initialize(FailureDetectorContext context)	Initialized this failure detector with the runtime context.
boolean	isAlive(ClusterAddress node)	Returns true if cluster node at the specified address is known to be alive.
boolean	isMonitored(ClusterAddress node)	Returns true if the specified node is in the list of monitored() nodes.
DefaultFailureDetectorJmx	jmx()	Creates a JMX object.
List<ClusterAddress>	monitored()	Returns addresses of all nodes that are monitored by this failure detector.
void	onConnectFailure(ClusterAddress node)	Notifies this failure detector upon failure while trying to connect to a remote node.
void	onHeartbeatReply(ClusterAddress node)	Notifies this failure detector on heartbeat reply message from a remote node.
boolean	onHeartbeatRequest(ClusterAddress from)	Notifies this failure detector on heartbeat request message form a remote node.
void	report(ConfigReporter report)	Report configuration.
void	terminate()	Terminates this failure detector.
String	toString()
void	update(Set<ClusterAddress> nodes)	Updates this failure detector with the latest information about all known cluster nodes addresses (including local node address).

Methods inherited from class java.lang.Object

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

Constructor Detail

DefaultFailureDetector

public DefaultFailureDetector()

Constructs a new instance with the default configuration options.

Please see the documentation of DefaultFailureDetectorConfig for information about default values.

DefaultFailureDetector

public DefaultFailureDetector(DefaultFailureDetectorConfig cfg)

Constructs a new instance using the specified configuration.

Parameters:

cfg - Configuration.

Method Detail

report

public void report(ConfigReporter report)

Description copied from interface: ConfigReportSupport

Report configuration.

Specified by:

report in interface ConfigReportSupport

Parameters:

report - Reporter.

initialize

public void initialize(FailureDetectorContext context)
                throws HekateException

Description copied from interface: FailureDetector

Initialized this failure detector with the runtime context.

Specified by:

initialize in interface FailureDetector

Parameters:

context - Context.

Throws:

HekateException - If this failure detector couldn't be initialized.

heartbeatInterval

public long heartbeatInterval()

Description copied from interface: FailureDetector

Returns the time interval in milliseconds between heartbeat sending rounds (see FailureDetector.heartbeatTick()).

If the returned value if less than or equals to zero then health monitoring will be completely disabled and FailureDetector.heartbeatTick()/FailureDetector.isAlive(ClusterAddress) methods will never be called.

Specified by:

heartbeatInterval in interface FailureDetector

Returns:

Time interval in milliseconds between heartbeat sending rounds.

failureQuorum

public int failureQuorum()

Description copied from interface: FailureDetector

Return the amount of nodes that should agree on some particular node failure before removing such node from the cluster.

The value of this parameter is expected to be greater than or equals to 1. If values is less then 1 then it will be automatically adjusted to 1.

Specified by:

failureQuorum in interface FailureDetector

Returns:

Amount of nodes that should agree on some particular node failure before removing such node from the cluster.

terminate

public void terminate()

Description copied from interface: FailureDetector

Terminates this failure detector.

Specified by:

terminate in interface FailureDetector

isAlive

public boolean isAlive(ClusterAddress node)

Description copied from interface: FailureDetector

Returns true if cluster node at the specified address is known to be alive. Returns false if node is considered to be failed.

Specified by:

isAlive in interface FailureDetector

Parameters:

node - Node address.

Returns:

true if node is alive or false if node is considered to be failed.

heartbeatTick

public Collection<ClusterAddress> heartbeatTick()

Description copied from interface: FailureDetector

Runs a heartbeat tick and returns a set of cluster node addresses that should received a heartbeat request message.

The time interval between heartbeat ticks is controlled by FailureDetector.heartbeatInterval() method.

Specified by:

heartbeatTick in interface FailureDetector

Returns:

Set of cluster node addresses for heartbeat request message sending.

See Also:

FailureDetector.onHeartbeatRequest(ClusterAddress)

onHeartbeatRequest

public boolean onHeartbeatRequest(ClusterAddress from)

Description copied from interface: FailureDetector

Notifies this failure detector on heartbeat request message form a remote node. Returns a boolean flag indicating whether a heartbeat reply should be send (true) or heartbeat replies are not supported (false).

Specified by:

onHeartbeatRequest in interface FailureDetector

Parameters:

from - Address of the heartbeat request sender node.

Returns:

true if heartbeat reply should be send back to the requester.

See Also:

FailureDetector.heartbeatTick(), FailureDetector.onHeartbeatReply(ClusterAddress)

onHeartbeatReply

public void onHeartbeatReply(ClusterAddress node)

Description copied from interface: FailureDetector

Notifies this failure detector on heartbeat reply message from a remote node.

Specified by:

onHeartbeatReply in interface FailureDetector

Parameters:

node - Address of heartbeat reply sender node.

onConnectFailure

public void onConnectFailure(ClusterAddress node)

Description copied from interface: FailureDetector

Notifies this failure detector upon failure while trying to connect to a remote node.

Specified by:

onConnectFailure in interface FailureDetector

Parameters:

node - Address of a failed node.

update

public void update(Set<ClusterAddress> nodes)

Description copied from interface: FailureDetector

Updates this failure detector with the latest information about all known cluster nodes addresses (including local node address).

Note that the specified addresses set can include nodes that just started joining and are not within cluster service's topology.

Specified by:

update in interface FailureDetector

Parameters:

nodes - Cluster node addresses.

heartbeatLossThreshold

public int heartbeatLossThreshold()

Returns the amount of heartbeats that can be lost before considering a node failure.

Returns:

Amount of heartbeats that can be lost before considering a node failure.

See Also:

DefaultFailureDetectorConfig.setHeartbeatLossThreshold(int)

monitored

public List<ClusterAddress> monitored()

Returns addresses of all nodes that are monitored by this failure detector.

Returns:

Addresses of all nodes that are monitored by this failure detector.

isMonitored

public boolean isMonitored(ClusterAddress node)

Returns true if the specified node is in the list of monitored() nodes.

Parameters:

node - Node to check.

Returns:

true if the specified node is in the list of monitored() nodes.

jmx

public DefaultFailureDetectorJmx jmx()

Description copied from interface: JmxSupport

Creates a JMX object.

Specified by:

jmx in interface JmxSupport<DefaultFailureDetectorJmx>

Returns:

JMX object.

toString

public String toString()

Overrides:

toString in class Object