On Tuesday, November 11, 2014 4:53:09 PM UTC-5, Eddie Hartman wrote:
> This looks strange, David, I would recommend opening a PMR with our awesome L2 team.
Thanks for the look over Eddie! Yes, I opened the PMR at the same time I posted: 48133,L6Q,000
I just got off the phone with Jason Williams who was very helpful. Unfortunately, out troubleshooting process resulted in the problem resolving itself. We did two things:
1) Revert the JVM/JRE to the original version.
2) Remove the user accounts JAVA_HOME and the $JAVA_HOME/bin path from the user environment.
This resulted in the problem disappearing. What odd is, we next reversed the above actions one by one to determine which solved the problem:
1) Re-add the new JVM/JRE.
After this, the problem was still gone.
2) Add back the env variables when the problem existed.
Unfortunately, this didn't result in the problem recurring.
So... our conclusion is that there must have been some cached class files in the JVM that must have been flushed/rebuilt based on our actions.
IBM are closing the PMR and documenting this behavior.