Services send messages using channels to communicate with each other. An endpoint is a communications point for these services and is used to send and receive the messages sent through the channels. As a result, a node with no endpoints can communicate with other nodes in the same cluster, but not with clients.

The primary benefit for creating a node without endpoints in Red Hat JBoss Data Grid involves data replication.

A node without any endpoints cannot be accessed by the client directly. As a result, they are primarily used to replicate data from other nodes that can communicate with clients. The result is a node with a backup copy of the data that cannot be accessed by the client, which protects it from failure via an error sent by the client.

Red Hat JBoss Data Grid provides a sample configuration to configure a node without an endpoint:

A standard configuration, such as a standalone high availability configuration, can be changed into a configuration for a node with no endpoints by removing the endpoints subsystem from the configuration file.

Remove the XML element that begins with the following:

<subsystem xmlns="urn:infinispan:server:endpoint:8.0">

Removing the endpoints subsystem ensures endpoints are removed from the configuration and that clustering is not possible.