In the messaging subsystem there is an attribute jdbc-network-timeout with default value 20 seconds. However this parameter is passed to ActiveMQ Artemis as 20 000 000 milliseconds (20 000 seconds).
For reproducing the issue just configure JDBC persistence store and attach debugger with breakpoint on AbstractJDBCDriver::connect.
Blocker priority was set because it is regression against Wildfly 12.
The issue was hit with Artemis 2.5.0 with https://github.com/jmesnil/wildfly/tree/WFLY-9407_upgrade_artemis_2.4.0_with_prefix (commit 51dd8102f103ccb0470a3cfc8713d3f9bdb1b65d)