com.atlassian.confluence.cluster.hazelcast.shareddata

Class HazelcastSharedDataManager

java.lang.Object

com.atlassian.confluence.cluster.hazelcast.shareddata.HazelcastSharedDataManager

All Implemented Interfaces:

SharedDataManager

@Internal
public class HazelcastSharedDataManager
extends Object
implements SharedDataManager

A Hazelcast implementation of SharedDataManager

Since:

5.7

Constructor Summary

Constructors

Constructor and Description
HazelcastSharedDataManager(com.google.common.base.Supplier<com.hazelcast.core.HazelcastInstance> hazelcastRef)

Method Summary

Modifier and Type	Method and Description
@NonNull SharedData	getSharedData(String name) Obtain a SharedDataManager for the given key.

Methods inherited from class java.lang.Object

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

Constructor Detail

HazelcastSharedDataManager

public HazelcastSharedDataManager(com.google.common.base.Supplier<com.hazelcast.core.HazelcastInstance> hazelcastRef)

Method Detail

getSharedData

public @NonNull SharedData getSharedData(String name)

Description copied from interface: SharedDataManager

Obtain a SharedDataManager for the given key. Data managed via that provider will be shared across all nodes of the cluster.

The returned shared data provider makes no consistency guarantees. On a single node cluster, or in a non-clustered installation, strong consistency can be assumed, and in a multi-node cluster a best-effort consistency will be provided. Some implementations may provide strong consistency, others may only provide weak and/or eventual consistency.

The shared data should be persistent as long as at least one node of the cluster remains running. If multiple cluster nodes fail simultaneously, then the data may be lost. Implementations may persist this data even if the whole cluster is restarted, but clients should not assume one way or the other.

Specified by:

getSharedData in interface SharedDataManager