We deployed infinispan cluster in openshift, 3 pods per data center. Sometimes (it can be running for days without problem but then suddenly it get locked and the lock never released until the pods get killed after fall the health check) we found the health check stopped responding and when looking to our Dynatrace locking track, we can seem some non blocking thread was locked when doing an ACL permission check. Our cache container authorization config something like this:

 
<security cache-size="1000" cache-timeout="1800000">
            <authorization>
                <cluster-role-mapper />
                <role name="app1" permissions="LISTEN READ WRITE EXEC MONITOR BULK_READ BULK_WRITE LIFECYCLE"/>
                <role name="app2" permissions="LISTEN READ WRITE EXEC MONITOR BULK_READ BULK_WRITE LIFECYCLE"/>
                <role name="app3" permissions="LISTEN READ WRITE EXEC MONITOR BULK_READ BULK_WRITE LIFECYCLE"/>
            </authorization>
        </security>
And we using properties-realm under security-realm
 <properties-realm groups-attribute="Roles">
                        <user-properties path="users.properties" relative-to="infinispan.server.config.path"
                                         plain-text="true"/>
                        <group-properties path="groups.properties" relative-to="infinispan.server.config.path"/>
/properties-realm>