Created attachment 802350 [details] Stack dumps of the VM including the identified deadlock at the end Description of problem: A deadlock was seen while running up a server on OpenShift, using the RTGov for ER3. The deadlock occurs deep in the ClassLoading, while The version of the JVM is java version "1.7.0_25" OpenJDK Runtime Environment (rhel-2.4.1.1.el6-i386) OpenJDK Server VM (build 24.0-b50, mixed mode) It looks like it is a bug in the JDK however I believe that weld may have a workaround, https://issues.jboss.org/browse/WELD-1169
The server reports itself as JBoss EAP 6.1.0.GA (AS 7.2.0.Final-redhat-8) and is an installation of the current SY + RTGov codebases.
This should be fixed by picking up EAP 6.1.1.
I double checked the EAP 6.1.1 release and, unfortunately, it does not contain the workaround for this JDK bug. The fix had been made in the Weld 2.x release stream whereas the EAP 6.1.1 release contains Weld-1.1.13. There is a separate issue for having this addressed in the 1.x stream, https://issues.jboss.org/browse/WELD-1296. I believe that this issue has only been seen once to date so it would appear that the window of opportunity is very small.
Looks like the problem has now been fixed in the jdk: http://bugs.java.com/bugdatabase/view_bug.do?bug_id=7122142
No longer a bug as it has been fixed in the latest JDK