Class DefaultClusterJoinManager
java.lang.Object
com.atlassian.confluence.impl.cluster.hazelcast.interceptor.authenticator.DefaultClusterJoinManager
- All Implemented Interfaces:
ClusterJoinManager
A
ClusterJoinManager
which first authenticates
new connections
Warning: Do not use AOP interceptors around this class!. Doing so can produce unexpected delays in join check processing, which can result in delays when nodes that are starting first attempt to join an existing cluster. That can result in nodes promoting themselves to full members and then clustering "later", defeating our configuration for nodes to join existing clusters as "lite" members (which can't hold partitioned data) until the application is fully started.
- Since:
- 7.17.3
-
Constructor Summary
ConstructorsConstructorDescriptionDefaultClusterJoinManager
(ClusterAuthenticator clusterAuthenticator, boolean enableNodeAuthentication) -
Method Summary
Modifier and TypeMethodDescriptionvoid
accept
(@NonNull ClusterJoinRequest request) The implementation of this method matches that ofconnect(ClusterJoinRequest)
but from the other side.void
connect
(@NonNull ClusterJoinRequest request) Attempt to join an existing cluster
-
Constructor Details
-
DefaultClusterJoinManager
public DefaultClusterJoinManager(ClusterAuthenticator clusterAuthenticator, boolean enableNodeAuthentication)
-
-
Method Details
-
accept
The implementation of this method matches that ofconnect(ClusterJoinRequest)
but from the other side.- Specified by:
accept
in interfaceClusterJoinManager
- Parameters:
request
- the join request- Throws:
IOException
- if the join negotiation failed, or a communication breakdown occurs- See Also:
-
connect
Description copied from interface:ClusterJoinManager
Attempt to join an existing cluster- Specified by:
connect
in interfaceClusterJoinManager
- Parameters:
request
- the join request- Throws:
IOException
- if the join negotiation failed, or a communication breakdown occurs
-