There are 2 currently unsupported tags replication-colocated and shared-store-colocated by which is possible create colocated HA topology for Artemis.

We need to document how to use them for this purpose and add warning that they're unsupported. Here is more information:

"Colocated HA topology configured using replication-colocated/shared-store-colocated" tags are planned to be supported in EAP 7.2 - RFE https://issues.jboss.org/browse/EAP7-605.

It's the new way how to configure colocated HA topology in Artemis. The basic idea is quite simple. There is Artemis cluster and by those tags we ask another (live) server in cluster to create backup for me. So If I'm live server and some other live server in cluster asks me to create backup for him then I will start backup server which is the same JVM as I.

What we support now and what we call "old" colocated HA topology is basically the same but it must be configured manually. We manually add new Artemis server configured as backup to another server which in Artemis cluster.

It's quite extensively described in documentation - https://access.redhat.com/documentation/en-us/red_hat_jboss_enterprise_application_platform/7.0/html-single/configuring_messaging/#colocated_shared_store

In configuration of messaging-activemq subsystem it looks like:
<subsystem xmlns="urn:jbossomain:messaging-activemq:1.1">
<server name="default">
<!-- configuration of live server -->
</server>
<server name="backup">
<!-- configuration of backup server for another live server in cluster -->
</server>
</subsystem>

The recommended way how to do it is to have always 2 EAP 7 server with 2 live/backup pairs like - EAP1(Live1/Backup2)<->EAP2(Live2/Backup1). If more servers are needed then add another 2 EAP 7 servers. This all work best with shared store. Current implementation of replicated journal requires special steps if both of the EAP 7 servers are stopped.