When running on an ORB that is shut down in another thread, the ORBRunner thread exit will whine:
WARN [com.arjuna.orbportability.logging.loggerI18N] (Thread-19) [com.arjuna.orbportability.OA.caughtexception] shutdownObject caught: org.omg.CORBA.OBJECT_NOT_EXIST: POA destroyed
We don't detect ORB shutdown internally, so do the cleanup unconditionally rather than only if the ORB is still running. Some resulting errors are already silently ignored. This one should be too - it's easier than adding lifecycle tracking to the orb portability layer.