On a six node (three live-backup pair) cluster, one of the brokers encountered the following exception and halted the system after 19 hours in operation. The live-backup pairs relied on data replication.

 
 
WARN [org.apache.activemq.artemis.utils.critical.CriticalMeasure] Component org.apache.activemq.artemis.core.persistence.impl.journal.JournalStorageManager is expired on path 0
ERROR [org.apache.activemq.artemis.core.server] AMQ224080: The server process will now be stopped, as component org.apache.activemq.artemis.core.persistence.impl.journal.JournalStorageManager@7f6c4db7 is not responsive
 
 

The logged information does not offer any insight into the underlying cause for the logged problem. It appears the JournalStorageManager throws the aforementioned exception should that encounter a timeout while accessing the persistent store. It would be best to increase the verbosity of the ERROR message since that would allow us to troubleshoot the problem without needing to enable more verbose logging at the TRACE level. In fact, it may not be possible to enable TRACE logging in production environments.